# Generated by `pnpm generate-openapi` from the Zod contract (app/api/v2/openapi/contract/). Do not edit by hand. openapi: 3.1.0 info: title: Artificial Analysis API version: "2.0" description: | Use the Artificial Analysis dataset inside your own products, research, analysis, and monitoring systems. The API provides structured model metadata, benchmark results, pricing, and performance data across language, image, video, music, and speech. ## License and usage Use of the API requires attribution across all tiers. When you display or share API data, credit [Artificial Analysis](https://artificialanalysis.ai/) as the source. A visible byline or footer link is sufficient. Use of the API is also subject to our [Terms of Use](https://artificialanalysis.ai/docs/legal/Terms-of-Use.pdf) and [Data Platform Terms](https://artificialanalysiscdn.com/legal/ProDataPlatformTerms.pdf). For redistribution rights or bespoke contract terms, [contact the team](https://artificialanalysis.ai/data-api#contact). ## Authentication All endpoints require an API key in the `x-api-key` header. ``` curl -H "x-api-key: YOUR_API_KEY" https://artificialanalysis.ai/api/v2/language/models ``` Create or copy a key from [API key management](https://artificialanalysis.ai/api-key-management-redirect). For Commercial access, [contact the team](https://artificialanalysis.ai/data-api#contact). ## Tiers | Tier | Access | Notes | | --- | --- | --- | | **Free** | The `/free` model-list endpoints — `/api/v2/language/models/free` and the 11 `/api/v2/media/.../models/free` siblings — return a Free-shape response with the public subset of fields | Standard non-`/free` routes are Pro+ and return `403` for Free API keys. Use the `/free` sibling instead | | **Pro** | Model-level access to the standard V2 endpoint set documented here, including LLM detail, pricing, richer performance data, and the Pro-shape model-list responses | This is the main paid tier for the new V2 API. Provider-level endpoints and performance over time are Commercial-only and return `403` for Pro API keys | | **Commercial** | Everything in Pro, plus provider-level language model data, performance over time (7, 30, and 90 day windows), and expanded data access negotiated on a per-organization basis. Examples include raw provider measurements, hardware data, or other ad-hoc additions | Commercial access is provisioned separately. [Contact the team](https://artificialanalysis.ai/data-api#contact) to discuss scope | ## Rate limits Rate limits use a fixed 24-hour window. The window does not use a rolling 24-hour period. The first request after the previous window ends starts a new window. The new window ends 24 hours after that request. The full quota resets when the window ends. The quota does not reset by hour or by request. For example, if you use 200 requests in one hour, they do not return separately 24 hours later. Each API key has a user and an organization. The rate limit applies to a shared scope, not to one API key. The scope covers one user in an organization or the full organization. All API keys in that scope share the quota. | Tier | Requests per 24-hour window | | --- | --- | | Free | 100 | | Pro | 500 | | Commercial | Custom. [Contact the team](https://artificialanalysis.ai/data-api#contact) | Use `X-RateLimit-Reset` as the exact reset time. ## Response headers Every authenticated response carries: - `X-AA-Tier` — the subscription tier that served the request (`free`, `pro`, or `commercial`). Mirrors the `tier` field in the response body. - `X-RateLimit-Limit`, `X-RateLimit-Remaining`, `X-RateLimit-Reset` — current rate-limit state. - `Retry-After` — included on `429` responses; seconds until the quota resets. ## Data conventions - Dates use **ISO 8601** format (for example, `2025-06-15` or `2025-06-15T14:30:00.000Z`) - Country codes use lowercased **ISO 3166-1 alpha-2** values (for example, `us`, `cn`, `gb`) - End-to-end response times assume **500 answer tokens** unless a field name states otherwise - Nulls mean "not measured" or "not applicable"; they never mean zero ## Intelligence Index versioning `artificial_analysis_intelligence_index` is a composite score, and what goes into it evolves. We version the index so you can interpret a score in context. Responses report the major.minor version that produced their scores in the `intelligence_index_version` field — treat that field, not this document, as the source of truth for the version a score belongs to. Full composition and version history on the [Intelligence Index methodology page](https://artificialanalysis.ai/methodology/intelligence-benchmarking). **How versions change:** - **Major version bumps** indicate a substantial change to the index definition or interpretation. Scores are best interpreted within the same major version. - **Minor version bumps** indicate methodology changes within the same major version. These may include changes to contributing evaluations, task subsets, graders, scoring/normalization, weightings, anchors, or evaluation refreshes. Scores may or may not change. - **Patch version bumps** indicate lower-significance methodology changes of the same kinds — for example upgrading an evaluation's grader or dataset version. Scores may change. They are not reflected in `intelligence_index_version`, which reports major.minor only. `artificial_analysis_coding_index` and `artificial_analysis_agentic_index` are always null: Artificial Analysis no longer publishes a Coding Index or an Agentic Index. The six Capability Indexes (`artificial_analysis_finance_and_accounting_index`, `artificial_analysis_strategy_and_ops_index`, `artificial_analysis_legal_index`, `artificial_analysis_healthcare_and_medical_index`, `artificial_analysis_engineering_index`, `artificial_analysis_economics_index`) are versioned separately from the Intelligence Index. Each is null for models missing any evaluation it depends on. Composition on the [Capability Indexes methodology page](https://artificialanalysis.ai/methodology/capability-indexes). servers: - url: https://artificialanalysis.ai description: Production security: - ApiKeyAuth: [] tags: - name: Language Models description: Evaluations, pricing, performance, and metadata for language models - name: Language Model Providers description: |- Capabilities, pricing, and performance for language model API providers. **Commercial only** — these endpoints require a Commercial API key. [Contact us](https://artificialanalysis.ai/data-api#contact) to discuss Commercial access. - name: Media - Image Arena description: AA-Image-T2I v2.1 and AA-Image-Editing v2.1 arena rankings for image generation models - name: Media - Video Arena description: AA-Video-T2V v2.1, AA-Video-T2V-Silent v2.1, AA-Video-I2V v2.1 and AA-Video-I2V-Silent v2.1 arena rankings for video generation models - name: Media - Speech description: Speech arena rankings, speech-to-speech data, and speech-to-text model data - name: Media - Music description: AA-Music-Instrumental v1.1 and AA-Music-Vocal v1.1 arena rankings for music generation models - name: Language Model Measurements description: | Raw measurement-level data for LLM performance benchmarking. **Commercial only** — this endpoint requires a Commercial API key with explicit provider-level grants. [Contact us](https://artificialanalysis.ai/data-api#contact) to discuss Commercial access. - name: CritPt Evaluation description: | Submit code solutions for grading against the CritPt v2 benchmark. Unlike the other endpoints in this API, this is **not** a data retrieval endpoint. It is a code evaluation service. You submit generated code for specific CritPt problems and receive grading results from the CritPt v2 evaluation server. The switch to the v2 grader does not change the endpoint or request and response schemas. **Access requires approval.** Grading is limited to approved labs and researchers — accounts without access receive a `403`. Access remains free. To request access, email [critpt@artificialanalysis.ai](mailto:critpt@artificialanalysis.ai) from your institutional address, including the email of the account that holds your API key and details of your research or development use. Use of this endpoint is subject to our [Terms of Use](https://artificialanalysis.ai/docs/legal/Terms-of-Use.pdf) and [Data Platform Terms](https://artificialanalysiscdn.com/legal/ProDataPlatformTerms.pdf). Rate limited to **10 requests per day** per user. Rate limit headers (`X-RateLimit-Limit`, `X-RateLimit-Remaining`, `X-RateLimit-Reset`, `Retry-After`) are included in every response. paths: /api/v2/language/models/free: get: operationId: getLanguageModelsFree summary: Get language models (Free tier) description: |- Free-tier sibling of `/api/v2/language/models`. Returns the Free-shape language model response: the public subset of fields. Includes headline indices (Intelligence, Coding, Agentic), the six Capability Indexes, median performance, Performance v3 headline means, and input/output pricing. Excludes Pro-only fields such as the full evaluation set, token counts, blended pricing, performance percentiles, time per task and per-workload Performance v3 metrics, context window, parameters, modalities, licensing, and provider detail. Accepts any valid API key (Free, Pro, or Commercial). Pro and Commercial keys also receive the Free-shape body when calling this URL; the `tier` field reflects the caller's actual tier. tags: - Language Models parameters: - $ref: "#/components/parameters/page" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/language/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/language/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/language/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/LLMModelsFreeResponse" examples: free_tier: summary: Free-shape response with `tier` field value: tier: free intelligence_index_version: 5 pagination: page: 1 page_size: 200 total_pages: 2 has_more: true data: - id: cd684ea4-b475-4269-b001-d469d06d8a7a name: GLM-5.3 (Max) slug: glm-5-3 release_date: 2026-08-18 model_creator: id: 67437eb6-7dc1-4e93-befd-22c8b8ec2065 name: Z AI evaluations: artificial_analysis_intelligence_index: 31.7 artificial_analysis_coding_index: null artificial_analysis_agentic_index: null artificial_analysis_finance_and_accounting_index: 42.1 artificial_analysis_strategy_and_ops_index: 47.4 artificial_analysis_legal_index: 35.8 artificial_analysis_healthcare_and_medical_index: 39.3 artificial_analysis_engineering_index: 37.6 artificial_analysis_economics_index: 39.3 artificial_analysis_intelligence_index_cost: total_cost: 3495.62 cost_per_task: total_cost: 5.0093 pricing: price_1m_input_tokens: 1.4 price_1m_output_tokens: 4.4 price_1m_cache_hit_tokens: 0.26 price_1m_cache_write_tokens: null performance: median_output_tokens_per_second: 80.07 median_time_to_first_token_seconds: 2.22 median_time_to_first_answer_token_seconds: 27.19 median_end_to_end_response_time_seconds: 33.44 performance_v3: mean_output_tokens_per_second: 62.36 mean_time_to_first_token_seconds: 3.45 mean_time_to_first_answer_token_seconds: 79.11 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/language/models: get: operationId: getLanguageModels summary: Get all language models description: |- Returns language model data including evaluations (headline indices, the six Capability Indexes, and per-benchmark scores), pricing, and performance metrics. Cyber Index scores come with `trusted_access`: `true` marks a security model that is not publicly available and has intentionally fewer safeguards. Detail-only fields (evaluation token counts, omniscience/openness breakdowns, providers) are available on `/api/v2/language/models/{slug}`. **Pro+ only.** Free integrations should use [`/api/v2/language/models/free`](#getLanguageModelsFree) instead. tags: - Language Models parameters: - $ref: "#/components/parameters/prompt_type" - $ref: "#/components/parameters/page" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/language/models" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/language/models", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/language/models", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/LLMModelsResponse" examples: pro_tier: summary: Pro tier example response value: tier: pro intelligence_index_version: 5 pagination: page: 1 page_size: 200 total_pages: 2 has_more: true data: - id: cd684ea4-b475-4269-b001-d469d06d8a7a name: GLM-5.3 (Max) slug: glm-5-3 release_date: 2026-08-18 model_creator: id: 67437eb6-7dc1-4e93-befd-22c8b8ec2065 name: Z AI country: cn reasoning_model: true trusted_access: false evaluations: artificial_analysis_intelligence_index: 31.7 artificial_analysis_coding_index: null artificial_analysis_agentic_index: null artificial_analysis_finance_and_accounting_index: 42.1 artificial_analysis_strategy_and_ops_index: 47.4 artificial_analysis_legal_index: 35.8 artificial_analysis_healthcare_and_medical_index: 39.3 artificial_analysis_engineering_index: 37.6 artificial_analysis_economics_index: 39.3 tau2_telecom: null tau_banking: 0.503 terminalbench_hard: null terminalbench_v2_1: 0.839 terminalbench_v4_0: 0.419 terminalbench_science: 0.095 aa_briefcase_elo: 1504 automationbench_aa: 0.622 gdp_pdf: 0.112 scicode: 0.59 aa_lcr: 0.797 aa_omniscience_index: 6.03 aa_omniscience_accuracy: 0.323 aa_omniscience_non_hallucination_rate: 0.611 ifbench: null hle: 0.423 gpqa_diamond: 0.917 critpt: 0.191 hle_diamond: 0.179 frontiercode_v1_1: 0.401 frontiercode_v2_0: 0.216 critpt_v2_hard: 0.335 gdpval_aa_elo: 1658 gdpval_aa_normalized: 0.579 mmmu_pro: null artificial_analysis_openness_index: 33.3 artificial_analysis_multilingual_index: null mlcr_overall: 0.172 apex_agents_aa: null aa_analyst_agent: 0.388 itbench_aa: 0.461 harvey_lab_aa_v1_1: 0.003 enterprise_ops_gym_aa: 0.364 healthbench: 0.545 healthbench_professional: 0.528 vials_aa: null critpt_v2: 0.574 artificial_analysis_cyber_index: 36.4 artificial_analysis_cyber_index_refusal_rate: 0 cwe_bench_aa: 0.558 cwe_bench_aa_refusal_rate: 0 deepsecbench_aa: 0.243 deepsecbench_aa_refusal_rate: 0 cybergym_e2e_aa: 0.29 cybergym_e2e_aa_refusal_rate: 0 artificial_analysis_intelligence_index_cost: total_cost: 3495.62 cost_per_task: total_cost: 5.0093 input_cost: 4.4857 reasoning_cost: 0.373 answer_cost: 0.1505 input_cost: 2901.32 reasoning_cost: 473.65 answer_cost: 120.65 artificial_analysis_intelligence_index_token_counts: input_tokens: 10296409355 answer_tokens: 27420388 output_tokens: 135067089 reasoning_tokens: 107646701 pricing: price_1m_blended_3_to_1: 2.15 price_1m_blended_7_to_2_to_1: 0.902 price_1m_input_tokens: 1.4 price_1m_output_tokens: 4.4 price_1m_cache_hit_tokens: 0.26 price_1m_cache_write_tokens: null performance: percentile_05_output_tokens_per_second: 35.32 quartile_25_output_tokens_per_second: 62.53 median_output_tokens_per_second: 80.07 quartile_75_output_tokens_per_second: 100.58 percentile_95_output_tokens_per_second: 117.05 percentile_05_time_to_first_token_seconds: 1.34 quartile_25_time_to_first_token_seconds: 1.83 median_time_to_first_token_seconds: 2.22 quartile_75_time_to_first_token_seconds: 2.94 percentile_95_time_to_first_token_seconds: 5.18 median_time_to_first_answer_token_seconds: 27.19 median_end_to_end_response_time_seconds: 33.44 performance_v3: mean_output_tokens_per_second: 62.36 time_per_task_seconds: 1907.94 mean_time_to_first_token_seconds: 3.45 mean_time_to_first_answer_token_seconds: 79.11 workloads: 1k: mean_time_to_first_token_seconds: 1.96 10k: mean_time_to_first_token_seconds: 2.03 100k: mean_time_to_first_token_seconds: 4.9 multi_turn: mean_time_to_first_token_seconds: 3.93 bucket_date: 2026-10-30 context_window_tokens: 1000000 parameters: total: 753 active: 40 modalities: input: text: true image: false video: false speech: false output: text: true image: false video: false speech: false licensing: is_open_weights: true huggingface_url: https://huggingface.co/zai-org/GLM-5.3 openrouter_api_id: z-ai/glm-5.3 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/language/models/{slug}: get: operationId: getLanguageModelBySlug summary: Get a single language model by slug description: |- Returns data for a single language model identified by its slug. Includes all pro tier list fields plus detail-only fields: evaluation token counts and omniscience/openness breakdowns. Pro and Commercial only. Free tier requests return a `403` subscription error. The per-provider breakdown (`providers` array) is **Commercial only** and omitted from Pro responses. [Contact us](https://artificialanalysis.ai/data-api#contact) to discuss Commercial access. tags: - Language Models parameters: - name: slug in: path required: true description: The model slug (e.g. "gemini-3-pro", "claude-opus-4-6", "deepseek-v3"). schema: type: string - $ref: "#/components/parameters/prompt_type" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/language/models/claude-sonnet-4?prompt_type=medium" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/language/models/claude-sonnet-4", headers={"x-api-key": "YOUR_API_KEY"}, params={"prompt_type": "medium"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/language/models/claude-sonnet-4?prompt_type=medium", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/LLMModelItemResponse" examples: commercial_tier: summary: Commercial tier detail example response (Pro responses omit the `providers` array) value: tier: commercial intelligence_index_version: 5 data: id: cd684ea4-b475-4269-b001-d469d06d8a7a name: GLM-5.3 (Max) slug: glm-5-3 release_date: 2026-08-18 model_creator: id: 67437eb6-7dc1-4e93-befd-22c8b8ec2065 name: Z AI country: cn reasoning_model: true trusted_access: false evaluations: artificial_analysis_intelligence_index: 31.7 artificial_analysis_coding_index: null artificial_analysis_agentic_index: null artificial_analysis_finance_and_accounting_index: 42.1 artificial_analysis_strategy_and_ops_index: 47.4 artificial_analysis_legal_index: 35.8 artificial_analysis_healthcare_and_medical_index: 39.3 artificial_analysis_engineering_index: 37.6 artificial_analysis_economics_index: 39.3 tau2_telecom: null tau_banking: 0.503 terminalbench_hard: null terminalbench_v2_1: 0.839 terminalbench_v4_0: 0.419 terminalbench_science: 0.095 aa_briefcase_elo: 1504 automationbench_aa: 0.622 gdp_pdf: 0.112 scicode: 0.59 aa_lcr: 0.797 aa_omniscience_index: 6.03 aa_omniscience_accuracy: 0.323 aa_omniscience_non_hallucination_rate: 0.611 ifbench: null hle: 0.423 gpqa_diamond: 0.917 critpt: 0.191 hle_diamond: 0.179 frontiercode_v1_1: 0.401 frontiercode_v2_0: 0.216 critpt_v2_hard: 0.335 gdpval_aa_elo: 1658 gdpval_aa_normalized: 0.579 mmmu_pro: null artificial_analysis_openness_index: 33.3 artificial_analysis_multilingual_index: null mlcr_overall: 0.172 apex_agents_aa: null aa_analyst_agent: 0.388 itbench_aa: 0.461 harvey_lab_aa_v1_1: 0.003 enterprise_ops_gym_aa: 0.364 healthbench: 0.545 healthbench_professional: 0.528 vials_aa: null critpt_v2: 0.574 artificial_analysis_cyber_index: 36.4 artificial_analysis_cyber_index_refusal_rate: 0 cwe_bench_aa: 0.558 cwe_bench_aa_refusal_rate: 0 deepsecbench_aa: 0.243 deepsecbench_aa_refusal_rate: 0 cybergym_e2e_aa: 0.29 cybergym_e2e_aa_refusal_rate: 0 artificial_analysis_intelligence_index_cost: total_cost: 3495.62 cost_per_task: total_cost: 5.0093 input_cost: 4.4857 reasoning_cost: 0.373 answer_cost: 0.1505 input_cost: 2901.32 reasoning_cost: 473.65 answer_cost: 120.65 artificial_analysis_intelligence_index_token_counts: input_tokens: 10296409355 answer_tokens: 27420388 output_tokens: 135067089 reasoning_tokens: 107646701 pricing: price_1m_blended_3_to_1: 2.15 price_1m_blended_7_to_2_to_1: 0.902 price_1m_input_tokens: 1.4 price_1m_output_tokens: 4.4 price_1m_cache_hit_tokens: 0.26 price_1m_cache_write_tokens: null performance: percentile_05_output_tokens_per_second: 35.32 quartile_25_output_tokens_per_second: 62.53 median_output_tokens_per_second: 80.07 quartile_75_output_tokens_per_second: 100.58 percentile_95_output_tokens_per_second: 117.05 percentile_05_time_to_first_token_seconds: 1.34 quartile_25_time_to_first_token_seconds: 1.83 median_time_to_first_token_seconds: 2.22 quartile_75_time_to_first_token_seconds: 2.94 percentile_95_time_to_first_token_seconds: 5.18 median_time_to_first_answer_token_seconds: 27.19 median_end_to_end_response_time_seconds: 33.44 performance_v3: mean_output_tokens_per_second: 62.36 time_per_task_seconds: 1907.94 mean_time_to_first_token_seconds: 3.45 mean_time_to_first_answer_token_seconds: 79.11 workloads: 1k: mean_time_to_first_token_seconds: 1.96 10k: mean_time_to_first_token_seconds: 2.03 100k: mean_time_to_first_token_seconds: 4.9 multi_turn: mean_time_to_first_token_seconds: 3.93 bucket_date: 2026-10-30 context_window_tokens: 1000000 parameters: total: 753 active: 40 modalities: input: text: true image: false video: false speech: false output: text: true image: false video: false speech: false licensing: is_open_weights: true huggingface_url: https://huggingface.co/zai-org/GLM-5.3 openrouter_api_id: z-ai/glm-5.3 evaluation_token_counts: hle: input_tokens: 694111 answer_tokens: 739298 output_tokens: 113978285 reasoning_tokens: 113238987 aa_lcr: input_tokens: 9731811 answer_tokens: 10191 output_tokens: 280574 reasoning_tokens: 270383 gpqa_diamond: input_tokens: 51246 answer_tokens: 29939 output_tokens: 1406523 reasoning_tokens: 1376584 tau_banking: input_tokens: 97164948 answer_tokens: 246937 output_tokens: 832923 reasoning_tokens: 585986 critpt: input_tokens: 620064 answer_tokens: 117126 output_tokens: 6996495 reasoning_tokens: 6879369 gdpval_aa: input_tokens: 1388705752 answer_tokens: 10535856 output_tokens: 20772998 reasoning_tokens: 10237142 scicode: input_tokens: 1020664 answer_tokens: 201454 output_tokens: 5044211 reasoning_tokens: 4842757 aa_omniscience: input_tokens: 377753 answer_tokens: 253325 output_tokens: 13870355 reasoning_tokens: 13617030 terminalbench_v2_1: input_tokens: 136692446 answer_tokens: 853063 output_tokens: 3551188 reasoning_tokens: 2698125 terminalbench_v4_0: input_tokens: 1717065779 answer_tokens: 3081614 output_tokens: 12841347 reasoning_tokens: 9759733 terminalbench_science: input_tokens: 3508627450 answer_tokens: 4943227 output_tokens: 18744947 reasoning_tokens: 13801720 aa_briefcase: input_tokens: 1627614045 answer_tokens: 7274345 output_tokens: 12406503 reasoning_tokens: 5132158 automationbench_aa: input_tokens: 656301228 answer_tokens: 6546304 output_tokens: 18224087 reasoning_tokens: 11677783 gdp_pdf: input_tokens: 3452755 answer_tokens: 162643 output_tokens: 2841710 reasoning_tokens: 2679067 hle_diamond: input_tokens: 266504 answer_tokens: 242177 output_tokens: 35780559 reasoning_tokens: 35538382 frontiercode_v1_1: input_tokens: 1419067666 answer_tokens: 2492536 output_tokens: 8344831 reasoning_tokens: 5852295 frontiercode_v2_0: input_tokens: 2683783478 answer_tokens: 4363236 output_tokens: 13978362 reasoning_tokens: 9615126 critpt_v2_hard: input_tokens: 98920363 answer_tokens: 553517 output_tokens: 6379219 reasoning_tokens: 5825702 aa_analyst_agent: input_tokens: 109474452 answer_tokens: 575076 output_tokens: 3058277 reasoning_tokens: 2483201 itbench_aa: input_tokens: 259691777 answer_tokens: 817281 output_tokens: 2792944 reasoning_tokens: 1975663 harvey_lab_aa_v1_1: input_tokens: 1691988614 answer_tokens: 12265548 output_tokens: 16553951 reasoning_tokens: 4288403 enterprise_ops_gym_aa: input_tokens: 1293941888 answer_tokens: 3219466 output_tokens: 52924711 reasoning_tokens: 49705245 healthbench: input_tokens: 825238 answer_tokens: 2508082 output_tokens: 13903947 reasoning_tokens: 11395865 healthbench_professional: input_tokens: 305622 answer_tokens: 584230 output_tokens: 3758230 reasoning_tokens: 3174000 critpt_v2: input_tokens: 139272171 answer_tokens: 916572 output_tokens: 9868665 reasoning_tokens: 8952093 aa_omniscience_breakdown: total: accuracy: 0.323 omniscience: 6.03 hallucination_rate: 0.389 law: accuracy: 0.153 omniscience: -6.89 hallucination_rate: 0.262 health: accuracy: 0.247 omniscience: -9.78 hallucination_rate: 0.457 business: accuracy: 0.236 omniscience: 5.11 hallucination_rate: 0.241 humanities_and_social_sciences: accuracy: 0.344 omniscience: 7.33 hallucination_rate: 0.414 science_engineering_and_mathematics: accuracy: 0.377 omniscience: 15 hallucination_rate: 0.364 software_engineering: accuracy: 0.505 omniscience: 18.33 hallucination_rate: 0.65 artificial_analysis_openness_index_breakdown: weights_access: 3 weights_license: 2 data_pretrain_access: 0 data_pretrain_license: 0 data_posttrain_access: 0 data_posttrain_license: 0 methodology_license: 0 methodology_disclosure: 1 artificial_analysis_openness_index: 33.3 providers: - id: b08f179e-4a69-4f35-b350-8f7144a8b7c8 name: Novita slug: novita endpoint_label: null pricing: price_1m_blended_3_to_1: 2.15 price_1m_blended_7_to_2_to_1: 0.902 price_1m_input_tokens: 1.4 price_1m_output_tokens: 4.4 price_1m_cache_hit_tokens: 0.26 price_1m_cache_write_tokens: null performance: percentile_05_output_tokens_per_second: 79.92 quartile_25_output_tokens_per_second: 107.88 median_output_tokens_per_second: 118.3 quartile_75_output_tokens_per_second: 136.1 percentile_95_output_tokens_per_second: 197.2 percentile_05_time_to_first_token_seconds: 1.59 quartile_25_time_to_first_token_seconds: 1.91 median_time_to_first_token_seconds: 2.3 quartile_75_time_to_first_token_seconds: 3.49 percentile_95_time_to_first_token_seconds: 4.51 median_time_to_first_answer_token_seconds: 19.2 median_end_to_end_response_time_seconds: 23.43 performance_v3: mean_output_tokens_per_second: 98.83 mean_time_to_first_token_seconds: 2.91 mean_time_to_first_answer_token_seconds: 57.53 endpoint_error_rate: 0 workloads: 1k: mean_output_tokens_per_second: 120.09 mean_time_to_first_token_seconds: 1.48 10k: mean_output_tokens_per_second: 128.5 mean_time_to_first_token_seconds: 2.38 100k: mean_output_tokens_per_second: 85.74 mean_time_to_first_token_seconds: 7.07 multi_turn: mean_output_tokens_per_second: 86.21 mean_time_to_first_token_seconds: 2.17 bucket_date: 2026-10-30 context_window_tokens: 1000000 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "404": $ref: "#/components/responses/NotFound" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/language/models/{slug}/performance: get: operationId: getLanguageModelPerformanceOverTime summary: Get performance over time for a single model description: |- Returns a time series of performance metrics for the specified model, broken down by API provider. Each provider endpoint that serves this model has its own data points within the requested window. - `series_v3`: Performance v3 for each period: end-to-end output speed, time to first token, time to first answer token and endpoint error rate, as means over the workloads measured ([methodology](https://artificialanalysis.ai/methodology/performance-benchmarking)). Data starts on September 15, 2026. - `series`: the older median metrics. Every metric returns `null` from December 10, 2026. See [Deprecations](https://artificialanalysis.ai/data-api/docs#deprecations). Use `window` to control how far back the data goes and `interval` to control the granularity (e.g. `window=90d&interval=weekly` returns ~13 weekly data points per provider). **Commercial only.** Free and Pro tier requests return a `403` subscription error. [Contact us](https://artificialanalysis.ai/data-api#contact) to discuss Commercial access. tags: - Language Models parameters: - name: slug in: path required: true description: The unique identifier for the model (e.g. `claude-sonnet-4`). Use the `/models` endpoint to discover available slugs. schema: type: string - name: window in: query description: Lookback period from today. Maximum 90 days for single-model endpoint. schema: type: string enum: - 7d - 30d - 90d default: 90d - $ref: "#/components/parameters/interval" - $ref: "#/components/parameters/prompt_type" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/language/models/claude-sonnet-4/performance?window=90d&interval=weekly" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/language/models/claude-sonnet-4/performance", headers={"x-api-key": "YOUR_API_KEY"}, params={"window": "90d", "interval": "weekly"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/language/models/claude-sonnet-4/performance?window=90d&interval=weekly", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ModelPerformanceOverTimeResponse" "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/CommercialAccessRequired" "404": $ref: "#/components/responses/NotFound" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/language/providers: get: operationId: getLanguageProviders summary: Get all language model API providers description: |- Returns data for all API providers including the models they offer, pricing, and median performance metrics. **Commercial only.** Free and Pro tier requests return a `403` subscription error. [Contact us](https://artificialanalysis.ai/data-api#contact) to discuss Commercial access. tags: - Language Model Providers parameters: - $ref: "#/components/parameters/page" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/language/providers" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/language/providers", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/language/providers", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/LLMProvidersResponse" examples: commercial_tier: summary: Commercial tier example response value: tier: commercial pagination: page: 1 page_size: 200 total_pages: 1 has_more: false data: - id: b08f179e-4a69-4f35-b350-8f7144a8b7c8 name: Novita slug: novita country: sg models: - id: cd684ea4-b475-4269-b001-d469d06d8a7a name: GLM-5.3 (Max) slug: glm-5-3 endpoint_label: null pricing: price_1m_blended_3_to_1: 2.15 price_1m_blended_7_to_2_to_1: 0.902 price_1m_input_tokens: 1.4 price_1m_output_tokens: 4.4 price_1m_cache_hit_tokens: 0.26 price_1m_cache_write_tokens: null performance: median_output_tokens_per_second: 113.67 median_time_to_first_token_seconds: 1.97 median_time_to_first_answer_token_seconds: 19.56 median_end_to_end_response_time_seconds: 23.96 performance_v3: mean_output_tokens_per_second: 98.83 mean_time_to_first_token_seconds: 2.91 mean_time_to_first_answer_token_seconds: 57.53 endpoint_error_rate: 0 bucket_date: 2026-10-30 context_window_tokens: 1000000 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/CommercialAccessRequired" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/language/providers/{slug}: get: operationId: getLanguageProviderBySlug summary: Get a single language model provider by slug description: |- Returns data for a single API provider identified by its slug. Includes all fields from the providers list plus full performance percentile and quartile data for each model. **Commercial only.** Free and Pro tier requests return a `403` subscription error. [Contact us](https://artificialanalysis.ai/data-api#contact) to discuss Commercial access. tags: - Language Model Providers parameters: - name: slug in: path required: true description: The provider slug (e.g. "openai", "fireworks"). schema: type: string - $ref: "#/components/parameters/prompt_type" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/language/providers/fireworks?prompt_type=medium" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/language/providers/fireworks", headers={"x-api-key": "YOUR_API_KEY"}, params={"prompt_type": "medium"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/language/providers/fireworks?prompt_type=medium", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/LLMProviderItemResponse" "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/CommercialAccessRequired" "404": $ref: "#/components/responses/NotFound" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/language/providers/{slug}/performance: get: operationId: getLanguageProviderPerformanceOverTime summary: Get performance over time for a single provider description: |- Returns performance time series for each model offered by the specified provider, looking back from today by the specified window. - `series_v3`: Performance v3 for each period: end-to-end output speed, time to first token, time to first answer token and endpoint error rate, as means over the workloads measured ([methodology](https://artificialanalysis.ai/methodology/performance-benchmarking)). Data starts on September 15, 2026. - `series`: the older median metrics. Every metric returns `null` from December 10, 2026. See [Deprecations](https://artificialanalysis.ai/data-api/docs#deprecations). This is the inverse of the model performance endpoint — here you pick a provider and see how each of their models has performed over time. **Commercial only.** Free and Pro tier requests return a `403` subscription error. [Contact us](https://artificialanalysis.ai/data-api#contact) to discuss Commercial access. tags: - Language Model Providers parameters: - name: slug in: path required: true description: The unique identifier for the provider (e.g. `anthropic`). Use the `/providers` endpoint to discover available slugs. schema: type: string - $ref: "#/components/parameters/window" - $ref: "#/components/parameters/interval" - $ref: "#/components/parameters/prompt_type" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/language/providers/anthropic/performance?window=90d&interval=weekly" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/language/providers/anthropic/performance", headers={"x-api-key": "YOUR_API_KEY"}, params={"window": "90d", "interval": "weekly"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/language/providers/anthropic/performance?window=90d&interval=weekly", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ProviderPerformanceOverTimeResponse" "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/CommercialAccessRequired" "404": $ref: "#/components/responses/NotFound" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/language/providers/{slug}/measurements: get: operationId: getLanguageProviderMeasurements summary: Get raw measurements for a provider description: |- Returns measurement records for the specified provider, grouped by model. - `measurements_v3`: one record per Performance v3 run and workload, with median output speed, time to first token, time to first answer token and end-to-end time, plus sample counts. - `measurements`: the older per-request records. Every metric returns `null` from December 10, 2026. See [Deprecations](https://artificialanalysis.ai/data-api/docs#deprecations). **Commercial tier only.** Your API key must have explicit grants for the requested provider. If you are interested in accessing this endpoint, please [contact us](https://artificialanalysis.ai/data-api#contact) to discuss Commercial access. tags: - Language Model Measurements parameters: - name: slug in: path required: true description: The unique identifier for the provider (e.g. `anthropic`). Your Commercial API key must have grants for this provider. schema: type: string - $ref: "#/components/parameters/prompt_type" - name: start_date in: query description: |- Only return measurements recorded on or after this date (YYYY-MM-DD). Defaults to one month ago if omitted; dates earlier than one month ago are clamped to one month ago. schema: type: string format: date example: 2025-01-01 x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/language/providers/anthropic/measurements?prompt_type=medium" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/language/providers/anthropic/measurements", headers={"x-api-key": "YOUR_API_KEY"}, params={"prompt_type": "medium"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/language/providers/anthropic/measurements?prompt_type=medium", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/MeasurementsResponse" "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/CommercialAccessRequired" "404": $ref: "#/components/responses/NotFound" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/text-to-image/models/free: get: operationId: getTextToImageArenaFree summary: Get AA-Image-T2I v2.1 rankings (Free tier) description: |- Free-tier sibling of `/api/v2/media/text-to-image/models`. Returns the Free-shape arena response — model identity plus Elo and CI only. Accepts any valid API key (Free, Pro, or Commercial); Pro and Commercial keys also receive the Free-shape body (Pro-only fields omitted; the `tier` field reflects the caller's actual tier) when calling this URL. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Image Arena x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/text-to-image/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/text-to-image/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/text-to-image/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ImageArenaFreeResponse" examples: free: summary: Free-shape response with `tier` field value: tier: free data: - id: a1ee4d6f-d136-434b-bb1d-066fe5f9bf6f name: GPT Image 1.5 (high) slug: openai-gpt_image-1-5 model_creator: id: e67e56e3-15cd-43db-b679-da4660a69f41 name: OpenAI elo: 1266 ci_95: 11 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/text-to-image/models: get: operationId: getTextToImageArena summary: Get AA-Image-T2I v2.1 rankings description: |- Returns arena Elo rankings for AA-Image-T2I v2.1, the text-to-image benchmark. Each model includes identity, Elo and CI, plus rank, samples, pricing, release date, open_weights_url, and optional category breakdowns. **Pro+ only.** Free integrations should use [`/api/v2/media/text-to-image/models/free`](#getTextToImageArenaFree) instead. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Image Arena parameters: - $ref: "#/components/parameters/include_categories" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/text-to-image/models?include_categories=true" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/text-to-image/models", headers={"x-api-key": "YOUR_API_KEY"}, params={"include_categories": "true"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/text-to-image/models?include_categories=true", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ImageArenaResponse" examples: pro: summary: Pro response with categories value: tier: pro data: - id: a1ee4d6f-d136-434b-bb1d-066fe5f9bf6f name: GPT Image 1.5 (high) slug: openai-gpt_image-1-5 model_creator: id: e67e56e3-15cd-43db-b679-da4660a69f41 name: OpenAI elo: 1266 ci_95: 11 rank: 1 samples: 4650 release_date: 2025-12-16 price_per_1k_images: 133 open_weights_url: null categories: - label: Architecture & Real Estate group: use_case elo: 1273 ci_95: 18 samples: 1719 "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/image-editing/models/free: get: operationId: getImageEditingArenaFree summary: Get AA-Image-Editing v2.1 rankings (Free tier) description: |- Free-tier sibling of `/api/v2/media/image-editing/models`. Returns the Free-shape arena response — model identity plus Elo and CI only. Accepts any valid API key (Free, Pro, or Commercial); Pro and Commercial keys also receive the Free-shape body (Pro-only fields omitted; the `tier` field reflects the caller's actual tier) when calling this URL. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Image Arena x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/image-editing/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/image-editing/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/image-editing/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ImageArenaFreeResponse" examples: free: summary: Free-shape response with `tier` field value: tier: free data: - id: ae424391-721f-4df7-bf24-a8bef3c4c46c name: Riverflow 2.0 slug: sourceful_riverflow-2 model_creator: id: 71724fa8-622e-455f-aaa2-ef3c5a067f8b name: Sourceful elo: 1283 ci_95: 9 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/image-editing/models: get: operationId: getImageEditingArena summary: Get AA-Image-Editing v2.1 rankings description: |- Returns arena Elo rankings for AA-Image-Editing v2.1, the image editing benchmark. Each model includes identity, Elo and CI, plus rank, samples, pricing, release date, open_weights_url, and optional category breakdowns. **Pro+ only.** Free integrations should use [`/api/v2/media/image-editing/models/free`](#getImageEditingArenaFree) instead. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Image Arena parameters: - $ref: "#/components/parameters/include_categories" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/image-editing/models?include_categories=true" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/image-editing/models", headers={"x-api-key": "YOUR_API_KEY"}, params={"include_categories": "true"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/image-editing/models?include_categories=true", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ImageArenaResponse" examples: pro: summary: Pro response with categories value: tier: pro data: - id: ae424391-721f-4df7-bf24-a8bef3c4c46c name: Riverflow 2.0 slug: sourceful_riverflow-2 model_creator: id: 71724fa8-622e-455f-aaa2-ef3c5a067f8b name: Sourceful elo: 1283 ci_95: 9 rank: 1 samples: 4650 release_date: 2025-12-16 price_per_1k_images: 133 open_weights_url: null categories: - label: Identity Preserving Edit group: editing_action elo: 1273 ci_95: 18 samples: 1719 "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/text-to-video/models/free: get: operationId: getTextToVideoArenaFree summary: Get AA-Video-T2V-Silent v2.1 rankings (Free tier) description: |- Free-tier sibling of `/api/v2/media/text-to-video/models`. Returns the Free-shape arena response — model identity plus Elo and CI only. Accepts any valid API key (Free, Pro, or Commercial); Pro and Commercial keys also receive the Free-shape body (Pro-only fields omitted; the `tier` field reflects the caller's actual tier) when calling this URL. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Video Arena x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/text-to-video/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/text-to-video/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/text-to-video/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/VideoArenaFreeResponse" examples: free: summary: Free-shape response with `tier` field value: tier: free data: - id: 36db1f4d-a9f6-49d1-a74d-06bddbf2fe4f name: Dreamina Seedance 2.0 720p slug: seedance-2-0 model_creator: id: 2354746c-4775-4a06-b64d-0ba4137785b8 name: ByteDance Seed elo: 1269 ci_95: 8 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/text-to-video/models: get: operationId: getTextToVideoArena summary: Get AA-Video-T2V-Silent v2.1 rankings description: |- Returns arena Elo rankings for AA-Video-T2V-Silent v2.1, the text-to-video benchmark without audio. Each model includes identity, Elo and CI, plus rank, samples, pricing, release date, open_weights_url, and optional category breakdowns. **Pro+ only.** Free integrations should use [`/api/v2/media/text-to-video/models/free`](#getTextToVideoArenaFree) instead. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Video Arena parameters: - $ref: "#/components/parameters/include_categories" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/text-to-video/models?include_categories=true" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/text-to-video/models", headers={"x-api-key": "YOUR_API_KEY"}, params={"include_categories": "true"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/text-to-video/models?include_categories=true", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/VideoArenaResponse" examples: pro_tier: summary: Pro tier response with categories value: tier: pro data: - id: 36db1f4d-a9f6-49d1-a74d-06bddbf2fe4f name: Dreamina Seedance 2.0 720p slug: seedance-2-0 model_creator: id: 2354746c-4775-4a06-b64d-0ba4137785b8 name: ByteDance Seed elo: 1269 ci_95: 8 rank: 1 samples: 4271 release_date: 2025-06-01 price_per_minute: 0.45 open_weights_url: null categories: - label: Camera Control group: capability elo: 1284 ci_95: 19 samples: 812 "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/image-to-video/models/free: get: operationId: getImageToVideoArenaFree summary: Get AA-Video-I2V-Silent v2.1 rankings (Free tier) description: |- Free-tier sibling of `/api/v2/media/image-to-video/models`. Returns the Free-shape arena response — model identity plus Elo and CI only. Accepts any valid API key (Free, Pro, or Commercial); Pro and Commercial keys also receive the Free-shape body (Pro-only fields omitted; the `tier` field reflects the caller's actual tier) when calling this URL. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Video Arena x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/image-to-video/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/image-to-video/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/image-to-video/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/VideoArenaFreeResponse" examples: free: summary: Free-shape response with `tier` field value: tier: free data: - id: 61270a9b-9d2e-4875-810f-e81508bc5504 name: Kling 2.1 Master slug: kling-2-1-master model_creator: id: 62cc833b-f55a-4489-bcae-54806d4b04ff name: Kuaishou KlingAI elo: 1184 ci_95: 10 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/image-to-video/models: get: operationId: getImageToVideoArena summary: Get AA-Video-I2V-Silent v2.1 rankings description: |- Returns arena Elo rankings for AA-Video-I2V-Silent v2.1, the image-to-video benchmark without audio. Each model includes identity, Elo and CI, plus rank, samples, pricing, release date, open_weights_url, and optional category breakdowns. **Pro+ only.** Free integrations should use [`/api/v2/media/image-to-video/models/free`](#getImageToVideoArenaFree) instead. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Video Arena parameters: - $ref: "#/components/parameters/include_categories" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/image-to-video/models?include_categories=true" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/image-to-video/models", headers={"x-api-key": "YOUR_API_KEY"}, params={"include_categories": "true"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/image-to-video/models?include_categories=true", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/VideoArenaResponse" examples: pro_tier: summary: Pro tier response with categories value: tier: pro data: - id: 61270a9b-9d2e-4875-810f-e81508bc5504 name: Kling 2.1 Master slug: kling-2-1-master model_creator: id: 62cc833b-f55a-4489-bcae-54806d4b04ff name: Kuaishou KlingAI elo: 1184 ci_95: 10 rank: 1 samples: 1890 release_date: 2025-03-27 price_per_minute: 0.5 open_weights_url: null categories: - label: Photorealistic group: style elo: 1196 ci_95: 21 samples: 354 "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/text-to-video-audio/models/free: get: operationId: getTextToVideoAudioArenaFree summary: Get AA-Video-T2V v2.1 rankings (Free tier) description: |- Free-tier sibling of `/api/v2/media/text-to-video-audio/models`. Returns the Free-shape arena response — model identity plus Elo and CI only. Accepts any valid API key (Free, Pro, or Commercial); Pro and Commercial keys also receive the Free-shape body (Pro-only fields omitted; the `tier` field reflects the caller's actual tier) when calling this URL. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Video Arena x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/text-to-video-audio/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/text-to-video-audio/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/text-to-video-audio/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/VideoArenaFreeResponse" examples: free: summary: Free-shape response with `tier` field value: tier: free data: - id: 5ad8e0d8-d9bf-4a55-904a-f98e7a7a4b2d name: Veo 3 Fast slug: veo-3-fast model_creator: id: ad0d7c6b-6b44-458a-a95f-9d0cb3d8f8e2 name: Google elo: 1221 ci_95: 14 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/text-to-video-audio/models: get: operationId: getTextToVideoAudioArena summary: Get AA-Video-T2V v2.1 rankings description: |- Returns arena Elo rankings for AA-Video-T2V v2.1, the text-to-video benchmark for models that generate video with audio output. Each model includes identity, Elo and CI, plus rank, samples, pricing, release date, open_weights_url, and optional category breakdowns. **Pro+ only.** Free integrations should use [`/api/v2/media/text-to-video-audio/models/free`](#getTextToVideoAudioArenaFree) instead. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Video Arena parameters: - $ref: "#/components/parameters/include_categories" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/text-to-video-audio/models?include_categories=true" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/text-to-video-audio/models", headers={"x-api-key": "YOUR_API_KEY"}, params={"include_categories": "true"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/text-to-video-audio/models?include_categories=true", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/VideoArenaResponse" examples: pro_tier: summary: Pro tier response with categories value: tier: pro data: - id: 5ad8e0d8-d9bf-4a55-904a-f98e7a7a4b2d name: Veo 3 Fast slug: veo-3-fast model_creator: id: ad0d7c6b-6b44-458a-a95f-9d0cb3d8f8e2 name: Google elo: 1221 ci_95: 14 rank: 1 samples: 2334 release_date: 2025-05-20 price_per_minute: 0.75 open_weights_url: null categories: - label: Audio Synchronization group: capability elo: 1237 ci_95: 27 samples: 401 "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/image-to-video-audio/models/free: get: operationId: getImageToVideoAudioArenaFree summary: Get AA-Video-I2V v2.1 rankings (Free tier) description: |- Free-tier sibling of `/api/v2/media/image-to-video-audio/models`. Returns the Free-shape arena response — model identity plus Elo and CI only. Accepts any valid API key (Free, Pro, or Commercial); Pro and Commercial keys also receive the Free-shape body (Pro-only fields omitted; the `tier` field reflects the caller's actual tier) when calling this URL. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Video Arena x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/image-to-video-audio/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/image-to-video-audio/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/image-to-video-audio/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/VideoArenaFreeResponse" examples: free: summary: Free-shape response with `tier` field value: tier: free data: - id: 8b7cdcab-6f22-4d95-8d0f-f09db1d2233d name: PixVerse V4.5 Audio slug: pixverse-v4-5-audio model_creator: id: 1cb39f34-a0b6-4ca8-82fd-fec0e068eb7a name: PixVerse elo: 1163 ci_95: 17 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/image-to-video-audio/models: get: operationId: getImageToVideoAudioArena summary: Get AA-Video-I2V v2.1 rankings description: |- Returns arena Elo rankings for AA-Video-I2V v2.1, the image-to-video benchmark for models that generate video with audio output. Each model includes identity, Elo and CI, plus rank, samples, pricing, release date, open_weights_url, and optional category breakdowns. **Pro+ only.** Free integrations should use [`/api/v2/media/image-to-video-audio/models/free`](#getImageToVideoAudioArenaFree) instead. Only models that hold a leaderboard rank are returned: current models with at least 500 arena appearances. Models below the threshold and superseded versions are omitted rather than returned with a null rank. tags: - Media - Video Arena parameters: - $ref: "#/components/parameters/include_categories" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/image-to-video-audio/models?include_categories=true" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/image-to-video-audio/models", headers={"x-api-key": "YOUR_API_KEY"}, params={"include_categories": "true"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/image-to-video-audio/models?include_categories=true", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/VideoArenaResponse" examples: pro_tier: summary: Pro tier response with categories value: tier: pro data: - id: 8b7cdcab-6f22-4d95-8d0f-f09db1d2233d name: PixVerse V4.5 Audio slug: pixverse-v4-5-audio model_creator: id: 1cb39f34-a0b6-4ca8-82fd-fec0e068eb7a name: PixVerse elo: 1163 ci_95: 17 rank: 1 samples: 1487 release_date: 2025-02-14 price_per_minute: 0.6 open_weights_url: null categories: - label: Photorealistic group: style elo: 1178 ci_95: 34 samples: 188 "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/text-to-speech/models/free: get: operationId: getTextToSpeechArenaFree summary: Get text-to-speech arena rankings (Free tier) description: |- Free-tier sibling of `/api/v2/media/text-to-speech/models`. Returns the Free-shape arena response — model identity plus Elo and CI only. Accepts any valid API key (Free, Pro, or Commercial); Pro and Commercial keys also receive the Free-shape body (Pro-only fields omitted; the `tier` field reflects the caller's actual tier) when calling this URL. tags: - Media - Speech x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/text-to-speech/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/text-to-speech/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/text-to-speech/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/TextToSpeechFreeResponse" examples: free: summary: Free-shape response with `tier` field value: tier: free data: - id: 6e0eadf0-3bf9-4424-a563-90292f2a0afd name: Inworld TTS 1.5 Max slug: tts-1-5-max model_creator: id: 3aab6d2e-04b0-43ee-996f-d8ef76ac3232 name: Inworld elo: 1240 ci_95: 22 "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/text-to-speech/models: get: operationId: getTextToSpeechArena summary: Get text-to-speech arena rankings description: |- Returns arena Elo rankings for text-to-speech models. Each model includes identity, Elo and CI, plus samples, release date, and pricing. **Pro+ only.** Free integrations should use [`/api/v2/media/text-to-speech/models/free`](#getTextToSpeechArenaFree) instead. tags: - Media - Speech x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/text-to-speech/models" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/text-to-speech/models", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/text-to-speech/models", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/TextToSpeechResponse" examples: pro: summary: Pro response value: tier: pro data: - id: 6e0eadf0-3bf9-4424-a563-90292f2a0afd name: Inworld TTS 1.5 Max slug: tts-1-5-max model_creator: id: 3aab6d2e-04b0-43ee-996f-d8ef76ac3232 name: Inworld elo: 1240 ci_95: 22 samples: 1330 release_date: 2026-01-21 price_per_1m_characters: 10 "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/speech-to-speech/models/free: get: operationId: getSpeechToSpeechFree summary: Get speech-to-speech model data (Free tier) description: |- Free-tier sibling of `/api/v2/media/speech-to-speech/models`. Returns the Free-shape response — model identity plus BBA and FDB scores, with Pro-only per-provider pricing and performance data omitted. Accepts any valid API key (Free, Pro, or Commercial); Pro and Commercial keys also receive the Free-shape body (Pro-only fields omitted; the `tier` field reflects the caller's actual tier) when calling this URL. tags: - Media - Speech x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/speech-to-speech/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/speech-to-speech/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/speech-to-speech/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/SpeechToSpeechFreeResponse" examples: free: summary: Free-shape response with `tier` field value: tier: free data: - id: eead24bb-ad0b-45be-b0ab-7b16381c2daf name: GPT 4o audio chatcompletions slug: 4o-audio-chatcompletions model_creator: id: e67e56e3-15cd-43db-b679-da4660a69f41 name: OpenAI bba_score: 0.54 fdb_score: null tau_voice_score: 0.37 "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/speech-to-speech/models: get: operationId: getSpeechToSpeech summary: Get speech-to-speech model data description: |- Returns data for speech-to-speech models including quality metrics (BBA and FDB scores). It also includes per-provider pricing and performance data. **Pro+ only.** Free integrations should use [`/api/v2/media/speech-to-speech/models/free`](#getSpeechToSpeechFree) instead. tags: - Media - Speech x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/speech-to-speech/models" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/speech-to-speech/models", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/speech-to-speech/models", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/SpeechToSpeechResponse" examples: pro_tier: summary: Pro tier response value: tier: pro data: - id: eead24bb-ad0b-45be-b0ab-7b16381c2daf name: GPT 4o audio chatcompletions slug: 4o-audio-chatcompletions model_creator: id: e67e56e3-15cd-43db-b679-da4660a69f41 name: OpenAI bba_score: 0.54 fdb_score: null tau_voice_score: 0.37 providers: - id: 6047c38a-af71-4b47-9d66-d361753f71b4 name: OpenAI slug: openai price_per_hour_input: 3.6 price_per_hour_output: 14.4 time_to_first_audio_seconds: 3.38 "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/speech-to-text/models/free: get: operationId: getSpeechToTextModelsFree summary: Get speech-to-text models (Free tier) description: |- Free-tier sibling of `/api/v2/media/speech-to-text/models`. Returns the Free-shape response — model identity plus the overall WER index, with Pro-only benchmark breakdowns, open-weights status, and per-provider data omitted. Accepts any valid API key (Free, Pro, or Commercial); Pro and Commercial keys also receive the Free-shape body (Pro-only fields omitted; the `tier` field reflects the caller's actual tier) when calling this URL. tags: - Media - Speech x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/speech-to-text/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/speech-to-text/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/speech-to-text/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/SpeechToTextModelsFreeResponse" examples: free: summary: Free-shape response with `tier` field value: tier: free data: - id: 865b890e-c487-4e78-a404-d1af7d7ee910 name: GPT-4o Mini Transcribe, OpenAI model_creator: id: e67e56e3-15cd-43db-b679-da4660a69f41 name: OpenAI aa_wer_index: 0.045 "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/speech-to-text/models: get: operationId: getSpeechToTextModels summary: Get speech-to-text models description: |- Returns speech-to-text model data including quality metrics (word error rates). Each model includes identity, the overall WER index, benchmark-specific WER breakdowns, open weights status, and per-provider data. **Pro+ only.** Free integrations should use [`/api/v2/media/speech-to-text/models/free`](#getSpeechToTextModelsFree) instead. tags: - Media - Speech x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/speech-to-text/models" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/speech-to-text/models", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/speech-to-text/models", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/SpeechToTextModelsResponse" examples: pro: summary: Pro response value: tier: pro data: - id: 865b890e-c487-4e78-a404-d1af7d7ee910 name: GPT-4o Mini Transcribe, OpenAI model_creator: id: e67e56e3-15cd-43db-b679-da4660a69f41 name: OpenAI aa_wer_index: 0.045 aa_agenttalk: 0.037 voxpopuli_cleaned_aa: 0.024 earnings_22_cleaned_aa: 0.081 open_weights: false providers: - id: 6047c38a-af71-4b47-9d66-d361753f71b4 name: OpenAI slug: openai price_per_1k_minutes: 3 median_speed_factor: 48.36 aa_wer_index: 0.045 aa_agenttalk: 0.037 voxpopuli_cleaned_aa: 0.024 earnings_22_cleaned_aa: 0.081 "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/music/instrumental/models/free: get: operationId: getMusicInstrumentalArenaFree summary: Get AA-Music-Instrumental v1.1 rankings (Free tier) description: |- Free-tier sibling of `/api/v2/media/music/instrumental/models`. Returns the Free-shape arena response — model identity plus Elo and CI only. Accepts any valid API key (Free, Pro, or Commercial); Pro and Commercial keys also receive the Free-shape body (Pro-only fields omitted; the `tier` field reflects the caller's actual tier) when calling this URL. tags: - Media - Music x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/music/instrumental/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/music/instrumental/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/music/instrumental/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/MusicArenaFreeResponse" examples: free: summary: Free-shape response with `tier` field value: tier: free data: - id: 1d06bf78-0a16-4fdf-8029-663a694e3470 name: Suno V4.5 model_creator: id: 30bf1911-2cd9-4bce-a179-009b174196fc name: Suno elo: 1120 ci_95: 15 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/music/instrumental/models: get: operationId: getMusicInstrumentalArena summary: Get AA-Music-Instrumental v1.1 rankings description: |- Returns arena Elo rankings for AA-Music-Instrumental v1.1, the instrumental music generation benchmark. Each model includes identity, Elo and CI, plus samples and optional genre breakdowns. **Pro+ only.** Free integrations should use [`/api/v2/media/music/instrumental/models/free`](#getMusicInstrumentalArenaFree) instead. tags: - Media - Music parameters: - $ref: "#/components/parameters/include_genres" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/music/instrumental/models?include_genres=true" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/music/instrumental/models", headers={"x-api-key": "YOUR_API_KEY"}, params={"include_genres": "true"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/music/instrumental/models?include_genres=true", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/MusicArenaResponse" examples: pro_tier: summary: Pro response with genre breakdowns value: tier: pro data: - id: b5b7f9d3-85d4-47dc-9213-fe09c1ea5da1 name: Eleven Music model_creator: id: 8d057760-0a8b-460a-9625-6dd261bb325c name: ElevenLabs elo: 1064 ci_95: 14 samples: 2396 genres: - label: Latin elo: 1063 ci_95: 48 samples: 214 "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/music/with-vocals/models/free: get: operationId: getMusicWithVocalsArenaFree summary: Get AA-Music-Vocal v1.1 rankings (Free tier) description: |- Free-tier sibling of `/api/v2/media/music/with-vocals/models`. Returns the Free-shape arena response — model identity plus Elo and CI only. Accepts any valid API key (Free, Pro, or Commercial); Pro and Commercial keys also receive the Free-shape body (Pro-only fields omitted; the `tier` field reflects the caller's actual tier) when calling this URL. tags: - Media - Music x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/music/with-vocals/models/free" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/music/with-vocals/models/free", headers={"x-api-key": "YOUR_API_KEY"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/music/with-vocals/models/free", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/MusicArenaFreeResponse" examples: free: summary: Free-shape response with `tier` field value: tier: free data: - id: 1d06bf78-0a16-4fdf-8029-663a694e3470 name: Suno V4.5 model_creator: id: 30bf1911-2cd9-4bce-a179-009b174196fc name: Suno elo: 1080 ci_95: 12 "400": $ref: "#/components/responses/BadRequest" "401": $ref: "#/components/responses/Unauthorized" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/media/music/with-vocals/models: get: operationId: getMusicWithVocalsArena summary: Get AA-Music-Vocal v1.1 rankings description: |- Returns arena Elo rankings for AA-Music-Vocal v1.1, the music generation with vocals benchmark. Each model includes identity, Elo and CI, plus samples and optional genre breakdowns. **Pro+ only.** Free integrations should use [`/api/v2/media/music/with-vocals/models/free`](#getMusicWithVocalsArenaFree) instead. tags: - Media - Music parameters: - $ref: "#/components/parameters/include_genres" x-codeSamples: - lang: Shell label: cURL source: | curl "https://artificialanalysis.ai/api/v2/media/music/with-vocals/models?include_genres=true" \ -H "x-api-key: YOUR_API_KEY" - lang: Python label: Python source: | import requests response = requests.get( "https://artificialanalysis.ai/api/v2/media/music/with-vocals/models", headers={"x-api-key": "YOUR_API_KEY"}, params={"include_genres": "true"} ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/media/music/with-vocals/models?include_genres=true", { headers: { "x-api-key": "YOUR_API_KEY" } } ); const data = await response.json(); responses: "200": description: Successful response headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/MusicArenaResponse" examples: pro_tier: summary: Pro response with genre breakdowns value: tier: pro data: - id: 1d06bf78-0a16-4fdf-8029-663a694e3470 name: Suno V4.5 model_creator: id: 30bf1911-2cd9-4bce-a179-009b174196fc name: Suno elo: 1080 ci_95: 12 samples: 3428 genres: - label: Alternative / Indie elo: 1064 ci_95: 46 samples: 244 "401": $ref: "#/components/responses/Unauthorized" "403": $ref: "#/components/responses/Forbidden" "429": $ref: "#/components/responses/RateLimited" "500": $ref: "#/components/responses/InternalError" /api/v2/critpt/evaluate: post: operationId: evaluateCritPt summary: Submit code for CritPt v2 evaluation description: | Submit generated code solutions for grading against CritPt v2 benchmark problems. This endpoint differs from the rest of the API. It is a **code evaluation service**, not a data retrieval endpoint. It accepts one or more code submissions and returns grading results from the CritPt v2 evaluation server. The switch to the v2 grader does not change the endpoint or request and response schemas. **Access requires approval.** To protect benchmark integrity and ensure server capacity, grading is limited to approved labs and researchers — accounts without access receive a `403`. Access remains free. To request access, email [critpt@artificialanalysis.ai](mailto:critpt@artificialanalysis.ai) from your institutional address, including the email of the account that holds your API key and details of your research or development use. Use of this endpoint is subject to our [Terms of Use](https://artificialanalysis.ai/docs/legal/Terms-of-Use.pdf) and [Data Platform Terms](https://artificialanalysiscdn.com/legal/ProDataPlatformTerms.pdf). Rate limited to **10 requests per day** per user. Only successful gradings (`200`) count toward the limit — grader error responses (`400`/`422`/`503`/`5xx`) and grader timeouts are refunded. Rate limit information is returned via `X-RateLimit-*` response headers. tags: - CritPt Evaluation requestBody: required: true content: application/json: schema: type: object required: - submissions - batch_metadata properties: submissions: type: array description: One or more code submissions to evaluate. items: type: object required: - problem_id - generated_code - model - generation_config properties: problem_id: type: string description: The CritPt problem identifier. generated_code: type: string description: The generated code solution to evaluate. model: type: string description: The model that generated the code. generation_config: type: object description: Configuration used for code generation (passed through to the grader). additionalProperties: true messages: type: array description: Optional conversation messages that produced the code. items: type: object additionalProperties: true batch_metadata: type: object description: Metadata about the submission batch (passed through to the grader). additionalProperties: true example: submissions: - problem_id: prob_001 generated_code: |- import numpy as np def solve(): return np.array([1, 2, 3]) model: claude-sonnet-4-20250514 generation_config: temperature: 0 max_tokens: 4096 batch_metadata: run_id: my-eval-run-001 x-codeSamples: - lang: Shell label: cURL source: | curl -X POST "https://artificialanalysis.ai/api/v2/critpt/evaluate" \ -H "x-api-key: YOUR_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "submissions": [{ "problem_id": "prob_001", "generated_code": "import numpy as np\n\ndef solve():\n return np.array([1, 2, 3])", "model": "claude-sonnet-4-20250514", "generation_config": { "temperature": 0, "max_tokens": 4096 } }], "batch_metadata": { "run_id": "my-eval-run-001" } }' - lang: Python label: Python source: | import requests response = requests.post( "https://artificialanalysis.ai/api/v2/critpt/evaluate", headers={"x-api-key": "YOUR_API_KEY", "Content-Type": "application/json"}, json={ "submissions": [{ "problem_id": "prob_001", "generated_code": "import numpy as np\n\ndef solve():\n return np.array([1, 2, 3])", "model": "claude-sonnet-4-20250514", "generation_config": {"temperature": 0, "max_tokens": 4096}, }], "batch_metadata": {"run_id": "my-eval-run-001"}, } ) data = response.json() - lang: JavaScript label: JavaScript source: | const response = await fetch( "https://artificialanalysis.ai/api/v2/critpt/evaluate", { method: "POST", headers: { "x-api-key": "YOUR_API_KEY", "Content-Type": "application/json", }, body: JSON.stringify({ submissions: [{ problem_id: "prob_001", generated_code: "import numpy as np\n\ndef solve():\n return np.array([1, 2, 3])", model: "claude-sonnet-4-20250514", generation_config: { temperature: 0, max_tokens: 4096 }, }], batch_metadata: { run_id: "my-eval-run-001" }, }), } ); const data = await response.json(); responses: "200": description: Evaluation results from the CritPt v2 grading server. headers: X-RateLimit-Limit: schema: type: integer description: Maximum requests allowed in the current window. X-RateLimit-Remaining: schema: type: integer description: Requests remaining in the current window. X-RateLimit-Reset: schema: type: integer description: Unix timestamp at which the current rate limit window resets. content: application/json: schema: description: Grading results returned by the CritPt v2 evaluation server. The response shape is determined by the upstream grader. type: object additionalProperties: true "400": description: Invalid request body, or the grading server rejected the batch (e.g. submission count mismatch). Not counted toward the rate limit. content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: Invalid body "401": $ref: "#/components/responses/Unauthorized" "403": description: The account has not been granted CritPt grading access. Access is approved case by case for labs and researchers — request it via critpt@artificialanalysis.ai. Not counted toward the rate limit. content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: CritPt grading requires approved access. To protect benchmark integrity and ensure server capacity, the grading endpoint is limited to approved labs and researchers; access remains free. To request access, email critpt@artificialanalysis.ai from your institutional address, including the email of the account that holds your API key and details of your research or development use. "422": description: The grading server rejected the submission payload (e.g. required keys missing). Not counted toward the rate limit. content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: Required keys missing from payload "429": description: Rate limit exceeded (10 requests per day). headers: X-RateLimit-Limit: schema: type: integer X-RateLimit-Remaining: schema: type: integer X-RateLimit-Reset: schema: type: integer description: Unix timestamp at which the current rate limit window resets. Retry-After: schema: type: integer description: Seconds until the rate limit resets. content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: Rate limit exceeded "500": $ref: "#/components/responses/InternalError" "503": description: The grading server is busy — retry later. Not counted toward the rate limit. content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: Server busy "504": description: The grading server timed out. Not counted toward the rate limit. content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: CritPt grading server timed out components: securitySchemes: ApiKeyAuth: type: apiKey in: header name: x-api-key description: API key for authentication. Obtain from your Artificial Analysis account. parameters: prompt_type: name: prompt_type in: query deprecated: true description: | Ignored from December 10, 2026. The prompt type preset used for performance benchmarking. Controls how many input tokens and parallel queries are used in latency/throughput measurements. Applies to the `performance` fields only; `performance_v3` does not depend on it. | prompt_type | input_tokens | parallel_queries | | --- | --- | --- | | `medium` | 1,000 | 1 | | `long` (default) | 10,000 | 1 | | `100k` | 100,000 | 1 | | `vision_single_image` | 1,000 | 1 | | `medium_parallel` | 1,000 | 10 | schema: type: string enum: - medium - long - 100k - vision_single_image - medium_parallel default: long window: name: window in: query description: Lookback period from today. schema: type: string enum: - 7d - 30d - 90d default: 90d interval: name: interval in: query description: | Aggregation granularity. Auto-derived from window if omitted (7d/30d → daily, 90d → weekly). schema: type: string enum: - daily - weekly - monthly page: name: page in: query description: | Page number for paginated results (1-indexed). Defaults to 1. schema: type: integer minimum: 1 maximum: 10000 default: 1 include_categories: name: include_categories in: query description: Include per-category Elo breakdowns (Pro tier only). schema: type: boolean default: false include_genres: name: include_genres in: query description: Include per-genre Elo breakdowns (Pro tier only). schema: type: boolean default: false headers: RateLimitLimit: schema: type: integer description: Maximum requests allowed in the current window. RateLimitRemaining: schema: type: integer description: Requests remaining in the current window. RateLimitReset: schema: type: integer description: Unix timestamp when the fixed 24-hour window ends and the full quota resets. Tier: schema: type: string enum: - free - pro - commercial description: | Subscription tier that served the request. Present on every authenticated response (success and error), and mirrors the `tier` field in successful response bodies. RetryAfter: schema: type: integer description: Seconds to wait before retrying. Present on `429` responses. responses: BadRequest: description: Invalid query parameters headers: X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: "Invalid prompt_type. Valid options: medium, long, 100k, vision_single_image, medium_parallel" Unauthorized: description: Missing or invalid API key content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: API key is required NotFound: description: Resource not found headers: X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: "Model not found: nonexistent-slug" Forbidden: description: Free tier requesting a pro-only feature headers: X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: Language models list requires a Pro subscription CommercialAccessRequired: description: Commercial access required — your API key does not have grants for this provider. headers: X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: Commercial access required RateLimited: description: Rate limited headers: X-RateLimit-Limit: $ref: "#/components/headers/RateLimitLimit" X-RateLimit-Remaining: $ref: "#/components/headers/RateLimitRemaining" X-RateLimit-Reset: $ref: "#/components/headers/RateLimitReset" Retry-After: $ref: "#/components/headers/RetryAfter" X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: Rate limit exceeded InternalError: description: Server error headers: X-AA-Tier: $ref: "#/components/headers/Tier" content: application/json: schema: $ref: "#/components/schemas/ErrorResponse" example: error: Internal Server Error schemas: ErrorResponse: type: object properties: error: type: string details: type: object description: Additional error details (e.g. validation errors). Only present on some error responses. required: - error PaginationMeta: type: object properties: page: type: integer description: Current page number (1-indexed). page_size: type: integer description: Number of items per page. total_pages: type: integer description: Total number of pages. has_more: type: boolean description: Whether there are more pages after the current one. required: - page - page_size - total_pages - has_more LLMModelsResponse: type: object properties: tier: type: string enum: - pro - commercial intelligence_index_version: $ref: "#/components/schemas/IntelligenceIndexVersion" pagination: $ref: "#/components/schemas/PaginationMeta" data: type: array items: $ref: "#/components/schemas/ModelDataExternal" required: - tier - intelligence_index_version - pagination - data LLMModelsFreeResponse: type: object properties: tier: type: string enum: - free - pro - commercial intelligence_index_version: $ref: "#/components/schemas/IntelligenceIndexVersion" pagination: $ref: "#/components/schemas/PaginationMeta" data: type: array items: $ref: "#/components/schemas/FreeModelData" required: - tier - intelligence_index_version - pagination - data LLMModelItemResponse: type: object properties: tier: type: string enum: - pro - commercial intelligence_index_version: $ref: "#/components/schemas/IntelligenceIndexVersion" data: $ref: "#/components/schemas/ModelDataExternal" required: - tier - intelligence_index_version - data IntelligenceIndexVersion: type: number description: Version of the Artificial Analysis Intelligence Index the `artificial_analysis_intelligence_index` scores in this response belong to, as a number with the `v` dropped (e.g. `5` for v5.0). Being a number, it carries major.minor only — patch versions are not reflected here. See the Intelligence Index versioning section for what major, minor, and patch changes mean. FreeModelData: type: object properties: id: type: string name: type: string slug: type: string release_date: description: ISO date string (e.g. "2024-05-13"). type: - string - "null" model_creator: oneOf: - $ref: "#/components/schemas/FreeModelCreator" - type: "null" evaluations: $ref: "#/components/schemas/FreeEvaluations" artificial_analysis_intelligence_index_cost: oneOf: - $ref: "#/components/schemas/FreeIntelligenceIndexCost" - type: "null" description: Pre-calculated cost to run the Intelligence Index evaluation. Free tier exposes total cost plus `cost_per_task.total_cost`. pricing: $ref: "#/components/schemas/FreePricing" description: Input and output token pricing. Pro tier adds blended pricing fields. performance: $ref: "#/components/schemas/FreeModelPerformance" description: Every field returns `null` from December 10, 2026. Use `performance_v3` ([methodology](https://artificialanalysis.ai/methodology/performance-benchmarking)). Median performance metrics across all providers serving this model. Pro tier adds percentiles and quartiles. performance_v3: oneOf: - $ref: "#/components/schemas/FreeModelPerformanceV3" - type: "null" description: Performance v3 headline means. Pro tier adds time per task, per-workload metrics and the window date. Null when the model has no measurements in the latest window. required: - id - name - slug - release_date - model_creator - evaluations - artificial_analysis_intelligence_index_cost - pricing - performance - performance_v3 description: Free-tier shape returned by `/api/v2/language/models/free`. Includes headline indices, median performance, Performance v3 headline means, and input/output pricing. Pro and Commercial keys also receive this shape when calling the `/free` endpoint. FreeModelCreator: type: object properties: id: type: string name: type: string required: - id - name FreeEvaluations: type: object properties: artificial_analysis_intelligence_index: description: AA Intelligence Index composite score. type: - number - "null" artificial_analysis_coding_index: description: "Always null: Artificial Analysis no longer publishes a Coding Index." deprecated: true type: - number - "null" artificial_analysis_agentic_index: description: "Always null: Artificial Analysis no longer publishes an Agentic Index." deprecated: true type: - number - "null" artificial_analysis_finance_and_accounting_index: description: AA Finance & Accounting Index capability score. Null when the model lacks a required evaluation. type: - number - "null" artificial_analysis_strategy_and_ops_index: description: AA Strategy & Ops Index capability score. Null when the model lacks a required evaluation. type: - number - "null" artificial_analysis_legal_index: description: AA Legal Index capability score. Null when the model lacks a required evaluation. type: - number - "null" artificial_analysis_healthcare_and_medical_index: description: AA Healthcare & Medical Index capability score. Null when the model lacks a required evaluation. type: - number - "null" artificial_analysis_engineering_index: description: AA Engineering Index capability score. Null when the model lacks a required evaluation. type: - number - "null" artificial_analysis_economics_index: description: AA Economics Index capability score. Null when the model lacks a required evaluation. type: - number - "null" required: - artificial_analysis_intelligence_index - artificial_analysis_coding_index - artificial_analysis_agentic_index - artificial_analysis_finance_and_accounting_index - artificial_analysis_strategy_and_ops_index - artificial_analysis_legal_index - artificial_analysis_healthcare_and_medical_index - artificial_analysis_engineering_index - artificial_analysis_economics_index description: Composite and capability Artificial Analysis indices. Pro tier adds per-benchmark scores. FreeIntelligenceIndexCost: type: object properties: total_cost: type: number description: Total cost in USD to run the Intelligence Index evaluation. cost_per_task: oneOf: - $ref: "#/components/schemas/FreeIntelligenceIndexCostPerTask" - type: "null" description: Weighted average cost per Intelligence Index task, rounded to 4 decimal places. Null when token or pricing data is incomplete. required: - total_cost - cost_per_task FreeIntelligenceIndexCostPerTask: type: object properties: total_cost: type: number description: Weighted average cost in USD per Intelligence Index task, across all Intelligence Index evaluations. Rounded to 4 decimal places. required: - total_cost FreePricing: type: object properties: price_1m_input_tokens: description: Price per 1M input tokens. Null when pricing data is not available. type: - number - "null" price_1m_output_tokens: description: Price per 1M output tokens. Null when pricing data is not available. type: - number - "null" price_1m_cache_hit_tokens: description: Price per 1M cached input (cache hit / read) tokens. Null when the provider does not report cache pricing. type: - number - "null" price_1m_cache_write_tokens: description: Price per 1M cache write tokens. Null when the provider does not report cache pricing. type: - number - "null" required: - price_1m_input_tokens - price_1m_output_tokens - price_1m_cache_hit_tokens - price_1m_cache_write_tokens FreeModelPerformance: type: object properties: median_output_tokens_per_second: type: - number - "null" median_time_to_first_token_seconds: type: - number - "null" median_time_to_first_answer_token_seconds: type: - number - "null" median_end_to_end_response_time_seconds: type: - number - "null" required: - median_output_tokens_per_second - median_time_to_first_token_seconds - median_time_to_first_answer_token_seconds - median_end_to_end_response_time_seconds description: Every field returns `null` from December 10, 2026. Use `performance_v3` ([methodology](https://artificialanalysis.ai/methodology/performance-benchmarking)). Median performance metrics across all providers serving this model. Individual fields are null when no measurements are available for that metric. deprecated: true FreeModelPerformanceV3: type: object properties: mean_output_tokens_per_second: description: Headline end-to-end output speed. The performance endpoint's mean over the workloads measured. For a model without a measured performance endpoint, the median of its endpoints' means. type: - number - "null" mean_time_to_first_token_seconds: description: The performance endpoint's mean over the workloads measured. For a model without a measured performance endpoint, the median of its endpoints' means. type: - number - "null" mean_time_to_first_answer_token_seconds: description: The performance endpoint's mean over the workloads measured. For a model without a measured performance endpoint, the median of its endpoints' means. type: - number - "null" required: - mean_output_tokens_per_second - mean_time_to_first_token_seconds - mean_time_to_first_answer_token_seconds description: "Performance v3 headline means: end-to-end speed averaged over the single-request and multi-turn agentic workloads measured, trailing 5-day window. The metrics of the model's performance endpoint, else the median across its endpoints. Null when there are no measurements in the latest window." ModelPerformanceV3: type: object properties: mean_output_tokens_per_second: description: Headline end-to-end output speed. The performance endpoint's mean over the workloads measured. For a model without a measured performance endpoint, the median of its endpoints' means. type: - number - "null" time_per_task_seconds: description: "Seconds per Intelligence Index task: the weighted output tokens per task over the headline speed. Null when token counts or speed are missing." type: - number - "null" mean_time_to_first_token_seconds: description: The performance endpoint's mean over the workloads measured. For a model without a measured performance endpoint, the median of its endpoints' means. type: - number - "null" mean_time_to_first_answer_token_seconds: description: The performance endpoint's mean over the workloads measured. For a model without a measured performance endpoint, the median of its endpoints' means. type: - number - "null" workloads: $ref: "#/components/schemas/PerformanceV3ModelWorkloads" bucket_date: type: string description: Last day of the trailing 5-day window these values cover. ISO date string (e.g. "2026-10-30"). required: - mean_output_tokens_per_second - time_per_task_seconds - mean_time_to_first_token_seconds - mean_time_to_first_answer_token_seconds - workloads - bucket_date description: "Performance v3 block: end-to-end speed averaged over the single-request and multi-turn agentic workloads measured, trailing 5-day window. The metrics of the model's performance endpoint, else the median across its endpoints. Null when there are no measurements in the latest window." ProviderModelPerformanceV3: type: object properties: mean_output_tokens_per_second: description: Headline end-to-end output speed. Mean over the workloads measured. type: - number - "null" mean_time_to_first_token_seconds: description: Mean over the workloads measured. type: - number - "null" mean_time_to_first_answer_token_seconds: description: Mean over the workloads measured. type: - number - "null" endpoint_error_rate: description: Endpoint error samples over all samples across the workloads measured (0-1). Null when no workload measured an output speed. type: - number - "null" workloads: $ref: "#/components/schemas/PerformanceV3EndpointWorkloads" bucket_date: type: string description: Last day of the trailing 5-day window these values cover. ISO date string (e.g. "2026-10-30"). required: - mean_output_tokens_per_second - mean_time_to_first_token_seconds - mean_time_to_first_answer_token_seconds - endpoint_error_rate - workloads - bucket_date description: "Performance v3 block for one endpoint: end-to-end speed averaged over the single-request and multi-turn agentic workloads measured, trailing 5-day window. Null when the endpoint has no measurements in the latest window." ProviderModelPerformanceV3List: type: object properties: mean_output_tokens_per_second: description: Headline end-to-end output speed. Mean over the workloads measured. type: - number - "null" mean_time_to_first_token_seconds: description: Mean over the workloads measured. type: - number - "null" mean_time_to_first_answer_token_seconds: description: Mean over the workloads measured. type: - number - "null" endpoint_error_rate: description: Endpoint error samples over all samples across the workloads measured (0-1). Null when no workload measured an output speed. type: - number - "null" bucket_date: type: string description: Last day of the trailing 5-day window these values cover. ISO date string (e.g. "2026-10-30"). required: - mean_output_tokens_per_second - mean_time_to_first_token_seconds - mean_time_to_first_answer_token_seconds - endpoint_error_rate - bucket_date description: "Performance v3 headline returned on the providers list endpoint: end-to-end speed averaged over the single-request and multi-turn agentic workloads measured, trailing 5-day window. The provider detail endpoint adds per-workload figures. Null when the endpoint has no measurements in the latest window." PerformanceV3EndpointWorkloads: type: object properties: 1k: $ref: "#/components/schemas/PerformanceV3EndpointWorkload" 10k: $ref: "#/components/schemas/PerformanceV3EndpointWorkload" 100k: $ref: "#/components/schemas/PerformanceV3EndpointWorkload" multi_turn: $ref: "#/components/schemas/PerformanceV3EndpointWorkload" description: "Per-workload figures, as on the website: single requests at three prompt lengths (`1k`, `10k` and `100k` input tokens) and `multi_turn`, the mean of the multi-turn agentic workloads. For an endpoint, each single-request value is the mean of per-run medians over the trailing 5-day window. For a model without a measured performance endpoint, each value is the median across its endpoints. A workload is omitted when it has no measurements in the window." PerformanceV3EndpointWorkload: type: object properties: mean_output_tokens_per_second: description: "End-to-end output speed: output tokens over the full request time, including time to first token. For `multi_turn`, output tokens over the time the endpoint took to answer its turns, excluding the time tools take to run." type: - number - "null" mean_time_to_first_token_seconds: description: Seconds to the first token, reasoning or answer. For `multi_turn`, the mean over each session's turns. type: - number - "null" required: - mean_output_tokens_per_second - mean_time_to_first_token_seconds description: Metrics for one workload on one endpoint. PerformanceV3ModelWorkloads: type: object properties: 1k: $ref: "#/components/schemas/PerformanceV3ModelWorkload" 10k: $ref: "#/components/schemas/PerformanceV3ModelWorkload" 100k: $ref: "#/components/schemas/PerformanceV3ModelWorkload" multi_turn: $ref: "#/components/schemas/PerformanceV3ModelWorkload" description: "Per-workload figures, as on the website: single requests at three prompt lengths (`1k`, `10k` and `100k` input tokens) and `multi_turn`, the mean of the multi-turn agentic workloads. For an endpoint, each single-request value is the mean of per-run medians over the trailing 5-day window. For a model without a measured performance endpoint, each value is the median across its endpoints. A workload is omitted when it has no measurements in the window." PerformanceV3ModelWorkload: type: object properties: mean_time_to_first_token_seconds: description: Seconds to the first token, reasoning or answer. For `multi_turn`, the mean over each session's turns. type: - number - "null" required: - mean_time_to_first_token_seconds description: Metrics for one workload on the model's performance endpoint, or the median across its endpoints when it has none. ModelDataExternal: type: object properties: id: type: string name: type: string slug: type: string release_date: description: ISO date string (e.g. "2024-05-13"). type: - string - "null" model_creator: oneOf: - $ref: "#/components/schemas/ModelCreator" - type: "null" reasoning_model: type: boolean description: Pro tier only. Omitted on Free tier. trusted_access: type: boolean description: "True for a model that is not publicly available: a security model with intentionally fewer safeguards, offered only through its creator's trusted-access program. Its Cyber Index scores are not directly comparable with publicly available models. Pro tier only. Omitted on Free tier." evaluations: $ref: "#/components/schemas/Evaluations" artificial_analysis_intelligence_index_cost: oneOf: - $ref: "#/components/schemas/IntelligenceIndexCost" - type: "null" description: Pre-calculated cost to run the Intelligence Index evaluation. Free tier includes total_cost and cost_per_task.total_cost; Pro tier adds input_cost, reasoning_cost, answer_cost at both levels. artificial_analysis_intelligence_index_token_counts: oneOf: - $ref: "#/components/schemas/IntelligenceIndexTokenCounts" - type: "null" description: Pro tier only. Omitted on Free tier. pricing: $ref: "#/components/schemas/Pricing" description: Full pricing object including blended fields. Free tier returns input and output token prices only. See `FreePricing`. performance: $ref: "#/components/schemas/ModelPerformance" description: Every field returns `null` from December 10, 2026. Use `performance_v3` ([methodology](https://artificialanalysis.ai/methodology/performance-benchmarking)). Full performance object including quartiles and percentiles. Free tier returns medians only. See `FreeModelPerformance`. performance_v3: oneOf: - $ref: "#/components/schemas/ModelPerformanceV3" - type: "null" description: Full performance v3 block including time per task, per-workload metrics and the window date. Free tier returns the headline means only. See `FreeModelPerformanceV3`. Null when the model has no measurements in the latest window. context_window_tokens: description: Pro tier only. Omitted on Free tier. Null when not set for the model. type: - integer - "null" parameters: oneOf: - $ref: "#/components/schemas/Parameters" - type: "null" description: Pro tier only. Omitted on Free tier. modalities: $ref: "#/components/schemas/Modalities" description: Pro tier only. Omitted on Free tier. licensing: $ref: "#/components/schemas/Licensing" description: Pro tier only. Omitted on Free tier. huggingface_url: description: Pro tier only. Omitted on Free tier. Canonical Hugging Face or model weights URL when available. type: - string - "null" openrouter_api_id: description: Pro tier only. Omitted on Free tier. OpenRouter API model identifier when available. type: - string - "null" evaluation_token_counts: oneOf: - $ref: "#/components/schemas/EvaluationTokenCounts" - type: "null" description: Pro tier only. Detail endpoint only (`/api/v2/language/models/{slug}`). Omitted on Free tier and list endpoint. aa_omniscience_breakdown: oneOf: - $ref: "#/components/schemas/OmniscienceBreakdown" - type: "null" description: Pro tier only. Detail endpoint only (`/api/v2/language/models/{slug}`). Omitted on Free tier and list endpoint. artificial_analysis_openness_index_breakdown: $ref: "#/components/schemas/OpennessBreakdown" description: Pro tier only. Detail endpoint only (`/api/v2/language/models/{slug}`). Omitted on Free tier and list endpoint. providers: type: array items: $ref: "#/components/schemas/ModelProviderDataExternal" description: Commercial tier only. Detail endpoint only (`/api/v2/language/models/{slug}`). Omitted on Free and Pro tiers and on the list endpoint. required: - id - name - slug - release_date - model_creator - evaluations - artificial_analysis_intelligence_index_cost - performance_v3 ModelCreator: type: object properties: id: type: string name: type: string country: description: Pro tier only. Omitted on Free tier. ISO 3166-1 alpha-2 country code of the model creator. type: - string - "null" required: - id - name Evaluations: type: object properties: artificial_analysis_intelligence_index: description: AA Intelligence Index composite score. type: - number - "null" artificial_analysis_coding_index: description: "Always null: Artificial Analysis no longer publishes a Coding Index." deprecated: true type: - number - "null" artificial_analysis_agentic_index: description: "Always null: Artificial Analysis no longer publishes an Agentic Index." deprecated: true type: - number - "null" artificial_analysis_finance_and_accounting_index: description: AA Finance & Accounting Index capability score. Null when the model lacks a required evaluation. type: - number - "null" artificial_analysis_strategy_and_ops_index: description: AA Strategy & Ops Index capability score. Null when the model lacks a required evaluation. type: - number - "null" artificial_analysis_legal_index: description: AA Legal Index capability score. Null when the model lacks a required evaluation. type: - number - "null" artificial_analysis_healthcare_and_medical_index: description: AA Healthcare & Medical Index capability score. Null when the model lacks a required evaluation. type: - number - "null" artificial_analysis_engineering_index: description: AA Engineering Index capability score. Null when the model lacks a required evaluation. type: - number - "null" artificial_analysis_economics_index: description: AA Economics Index capability score. Null when the model lacks a required evaluation. type: - number - "null" tau2_telecom: deprecated: true description: τ²-Bench Telecom score. Pro tier only. Omitted on Free tier. type: - number - "null" tau_banking: deprecated: true description: τ³-Banking score. Pro tier only. Omitted on Free tier. type: - number - "null" terminalbench_hard: deprecated: true description: Terminal-Bench Hard score. Pro tier only. Omitted on Free tier. type: - number - "null" terminalbench_v2_1: deprecated: true description: Terminal-Bench 2.1 score. Pro tier only. Omitted on Free tier. type: - number - "null" terminalbench_v4_0: description: Terminal-Bench 4.0 score. Pro tier only. Omitted on Free tier. type: - number - "null" terminalbench_science: description: Terminal-Bench-Science 0.1 score. Pro tier only. Omitted on Free tier. type: - number - "null" aa_briefcase_elo: description: AA-Briefcase v1.1 Elo score. Pro tier only. Omitted on Free tier. type: - number - "null" automationbench_aa: description: AutomationBench-AA partial score. Pro tier only. Omitted on Free tier. type: - number - "null" gdp_pdf: description: GDP.pdf all-pass score. Pro tier only. Omitted on Free tier. type: - number - "null" scicode: description: SciCode score. Pro tier only. Omitted on Free tier. type: - number - "null" aa_lcr: description: AA-LCR score. Pro tier only. Omitted on Free tier. type: - number - "null" aa_omniscience_index: description: AA-Omniscience Net Score. Pro tier only. Omitted on Free tier. type: - number - "null" aa_omniscience_accuracy: description: Aggregate AA-Omniscience accuracy. Pro tier only. Omitted on Free tier. type: - number - "null" aa_omniscience_non_hallucination_rate: description: Aggregate AA-Omniscience non-hallucination score (1 - hallucination_rate). Pro tier only. Omitted on Free tier. type: - number - "null" ifbench: description: IFBench score. Pro tier only. Omitted on Free tier. type: - number - "null" hle: description: Humanity's Last Exam score. Pro tier only. Omitted on Free tier. type: - number - "null" gpqa_diamond: description: GPQA Diamond score. Pro tier only. Omitted on Free tier. type: - number - "null" critpt: description: CritPt score. Pro tier only. Omitted on Free tier. type: - number - "null" hle_diamond: description: HLE-Diamond score. Pro tier only. Omitted on Free tier. type: - number - "null" frontiercode_v1_1: description: FrontierCode 1.1 score. Pro tier only. Omitted on Free tier. type: - number - "null" frontiercode_v2_0: description: FrontierCode-AA 2.0 score. Pro tier only. Omitted on Free tier. type: - number - "null" critpt_v2_hard: description: CritPt v2 score on its 40-problem hard subset. Pro tier only. Omitted on Free tier. type: - number - "null" gdpval_aa_elo: description: GDPval raw Elo score. Pro tier only. Omitted on Free tier. type: - number - "null" gdpval_aa_normalized: description: GDPval score normalized to a 0-1 scale. Pro tier only. Omitted on Free tier. type: - number - "null" mmmu_pro: description: MMMU-Pro visual reasoning score. Pro tier only. Omitted on Free tier. type: - number - "null" artificial_analysis_openness_index: description: Openness Index score. Pro tier only. Omitted on Free tier. type: - number - "null" artificial_analysis_multilingual_index: description: Multilingual Index score. Pro tier only. Omitted on Free tier. type: - number - "null" mlcr_overall: description: MLCR-AA (Medical Long Context Reasoning) overall pass rate. Pro tier only. Omitted on Free tier. type: - number - "null" apex_agents_aa: description: "APEX-Agents-AA pass@1 success rate (0-1): share of tasks where the model fully satisfies the grading rubric. Pro tier only. Omitted on Free tier." type: - number - "null" aa_analyst_agent: description: "AA-AnalystAgent pass^5 score (0-1): share of tasks passed on all five attempts. Pro tier only. Omitted on Free tier." type: - number - "null" itbench_aa: description: ITBench-AA average precision at full recall on its SRE tasks, Kubernetes incident root-cause analysis (0-1). Pro tier only. Omitted on Free tier. type: - number - "null" harvey_lab_aa_v1_1: description: Harvey LAB-AA v1.1 hallucination-gated all-pass rate (0-1). Null when the run has no hallucination check. Pro tier only. Omitted on Free tier. type: - number - "null" enterprise_ops_gym_aa: description: EnterpriseOps-Gym-AA success rate (0-1). Pro tier only. Omitted on Free tier. type: - number - "null" healthbench: description: HealthBench length-adjusted score (0-1). Pro tier only. Omitted on Free tier. type: - number - "null" healthbench_professional: description: HealthBench Professional length-adjusted score (0-1). Pro tier only. Omitted on Free tier. type: - number - "null" vials_aa: description: "VIALS-AA pass^3 score (0-1): share of tasks passed on all three attempts. Pro tier only. Omitted on Free tier." type: - number - "null" critpt_v2: description: CritPt v2 score on the full 70-problem set (0-1). See `critpt_v2_hard` for the hard subset. Pro tier only. Omitted on Free tier. type: - number - "null" artificial_analysis_cyber_index: description: Artificial Analysis Cyber Index v1 (0-100), the mean of CWE-Bench-AA, DeepsecBench-AA and CyberGym-E2E-AA. Read with `trusted_access`. Pro tier only. Omitted on Free tier. type: - number - "null" artificial_analysis_cyber_index_refusal_rate: description: Mean of the three Cyber Index evaluations' refusal rates (0-1). Null unless all three report one. Pro tier only. Omitted on Free tier. type: - number - "null" cwe_bench_aa: description: CWE-Bench-AA vulnerability patching success, pass@1 (0-1). Pro tier only. Omitted on Free tier. type: - number - "null" cwe_bench_aa_refusal_rate: description: Share of CWE-Bench-AA tasks the model or its provider declined on safety grounds, as flagged by the provider's safety systems (0-1). Pro tier only. Omitted on Free tier. type: - number - "null" deepsecbench_aa: description: DeepsecBench-AA F2 score (0-1), weighting recall above precision. Pro tier only. Omitted on Free tier. type: - number - "null" deepsecbench_aa_refusal_rate: description: Share of DeepsecBench-AA tasks the model or its provider declined on safety grounds, as flagged by the provider's safety systems (0-1). Pro tier only. Omitted on Free tier. type: - number - "null" cybergym_e2e_aa: description: CyberGym-E2E-AA vulnerability discovery and patching success, pass@1 (0-1). Pro tier only. Omitted on Free tier. type: - number - "null" cybergym_e2e_aa_refusal_rate: description: Share of CyberGym-E2E-AA tasks the model or its provider declined on safety grounds, as flagged by the provider's safety systems (0-1). Pro tier only. Omitted on Free tier. type: - number - "null" required: - artificial_analysis_intelligence_index - artificial_analysis_coding_index - artificial_analysis_agentic_index - artificial_analysis_finance_and_accounting_index - artificial_analysis_strategy_and_ops_index - artificial_analysis_legal_index - artificial_analysis_healthcare_and_medical_index - artificial_analysis_engineering_index - artificial_analysis_economics_index Pricing: type: object properties: price_1m_blended_3_to_1: description: Blended price per 1M tokens (3:1 Input-Output ratio). Null when pricing data is not available. type: - number - "null" price_1m_blended_7_to_2_to_1: description: Blended price per 1M tokens (7:2:1 Cache-Input-Output ratio). Null when pricing data is not available. type: - number - "null" price_1m_input_tokens: description: Price per 1M input tokens. Null when pricing data is not available. type: - number - "null" price_1m_output_tokens: description: Price per 1M output tokens. Null when pricing data is not available. type: - number - "null" price_1m_cache_hit_tokens: description: Price per 1M cached input (cache hit / read) tokens. Null when the provider does not report cache pricing. type: - number - "null" price_1m_cache_write_tokens: description: Price per 1M cache write tokens. Null when the provider does not report cache pricing. type: - number - "null" required: - price_1m_blended_3_to_1 - price_1m_blended_7_to_2_to_1 - price_1m_input_tokens - price_1m_output_tokens - price_1m_cache_hit_tokens - price_1m_cache_write_tokens ModelPerformance: type: object properties: percentile_05_output_tokens_per_second: type: - number - "null" quartile_25_output_tokens_per_second: type: - number - "null" median_output_tokens_per_second: type: - number - "null" quartile_75_output_tokens_per_second: type: - number - "null" percentile_95_output_tokens_per_second: type: - number - "null" percentile_05_time_to_first_token_seconds: type: - number - "null" quartile_25_time_to_first_token_seconds: type: - number - "null" median_time_to_first_token_seconds: type: - number - "null" quartile_75_time_to_first_token_seconds: type: - number - "null" percentile_95_time_to_first_token_seconds: type: - number - "null" median_time_to_first_answer_token_seconds: type: - number - "null" median_end_to_end_response_time_seconds: type: - number - "null" required: - percentile_05_output_tokens_per_second - quartile_25_output_tokens_per_second - median_output_tokens_per_second - quartile_75_output_tokens_per_second - percentile_95_output_tokens_per_second - percentile_05_time_to_first_token_seconds - quartile_25_time_to_first_token_seconds - median_time_to_first_token_seconds - quartile_75_time_to_first_token_seconds - percentile_95_time_to_first_token_seconds - median_time_to_first_answer_token_seconds - median_end_to_end_response_time_seconds description: Every field returns `null` from December 10, 2026. Use `performance_v3` ([methodology](https://artificialanalysis.ai/methodology/performance-benchmarking)). Aggregated performance metrics across all providers serving this model. Individual fields are null when no measurements are available for that metric. deprecated: true Parameters: type: object properties: total: type: number description: Total model parameters (in billions). active: description: Active parameters per forward pass (relevant for MoE models). type: - number - "null" required: - total - active Modalities: type: object properties: input: $ref: "#/components/schemas/ModalitySet" output: $ref: "#/components/schemas/ModalitySet" required: - input - output ModalitySet: type: object properties: text: type: boolean image: type: boolean video: type: boolean speech: type: boolean required: - text - image - video - speech Licensing: type: object properties: is_open_weights: type: boolean required: - is_open_weights EvalTokenCounts: type: object properties: input_tokens: type: number answer_tokens: type: number output_tokens: type: number reasoning_tokens: type: number required: - input_tokens - answer_tokens - output_tokens - reasoning_tokens EvaluationTokenCounts: type: object properties: hle: $ref: "#/components/schemas/EvalTokenCounts" aa_lcr: $ref: "#/components/schemas/EvalTokenCounts" gpqa_diamond: $ref: "#/components/schemas/EvalTokenCounts" tau2_telecom: deprecated: true $ref: "#/components/schemas/EvalTokenCounts" tau_banking: deprecated: true $ref: "#/components/schemas/EvalTokenCounts" critpt: $ref: "#/components/schemas/EvalTokenCounts" gdpval_aa: $ref: "#/components/schemas/EvalTokenCounts" ifbench: $ref: "#/components/schemas/EvalTokenCounts" scicode: $ref: "#/components/schemas/EvalTokenCounts" mmmu_pro: $ref: "#/components/schemas/EvalTokenCounts" aa_omniscience: $ref: "#/components/schemas/EvalTokenCounts" terminalbench_hard: deprecated: true $ref: "#/components/schemas/EvalTokenCounts" terminalbench_v2_1: deprecated: true $ref: "#/components/schemas/EvalTokenCounts" terminalbench_v4_0: $ref: "#/components/schemas/EvalTokenCounts" terminalbench_science: $ref: "#/components/schemas/EvalTokenCounts" aa_briefcase: $ref: "#/components/schemas/EvalTokenCounts" automationbench_aa: $ref: "#/components/schemas/EvalTokenCounts" gdp_pdf: $ref: "#/components/schemas/EvalTokenCounts" hle_diamond: $ref: "#/components/schemas/EvalTokenCounts" frontiercode_v1_1: $ref: "#/components/schemas/EvalTokenCounts" frontiercode_v2_0: $ref: "#/components/schemas/EvalTokenCounts" critpt_v2_hard: $ref: "#/components/schemas/EvalTokenCounts" apex_agents_aa: $ref: "#/components/schemas/EvalTokenCounts" aa_analyst_agent: $ref: "#/components/schemas/EvalTokenCounts" itbench_aa: $ref: "#/components/schemas/EvalTokenCounts" harvey_lab_aa_v1_1: $ref: "#/components/schemas/EvalTokenCounts" enterprise_ops_gym_aa: $ref: "#/components/schemas/EvalTokenCounts" healthbench: $ref: "#/components/schemas/EvalTokenCounts" healthbench_professional: $ref: "#/components/schemas/EvalTokenCounts" vials_aa: $ref: "#/components/schemas/EvalTokenCounts" critpt_v2: $ref: "#/components/schemas/EvalTokenCounts" description: Per-evaluation token counts for published benchmarks. IntelligenceIndexCost: type: object properties: total_cost: type: number description: Total cost in USD to run the Intelligence Index evaluation. Available on all tiers. cost_per_task: oneOf: - $ref: "#/components/schemas/IntelligenceIndexCostPerTask" - type: "null" description: Weighted average cost per Intelligence Index task, rounded to 4 decimal places. Null when token or pricing data is incomplete. input_cost: description: Pro tier only. Omitted on Free tier. Null when not measured. type: - number - "null" reasoning_cost: description: Pro tier only. Omitted on Free tier. Null when not measured. type: - number - "null" answer_cost: description: Pro tier only. Omitted on Free tier. Null when not measured. type: - number - "null" required: - total_cost - cost_per_task IntelligenceIndexCostPerTask: type: object properties: total_cost: type: number description: Weighted average cost in USD per Intelligence Index task, across all Intelligence Index evaluations. Rounded to 4 decimal places. input_cost: description: Weighted average input cost in USD per Intelligence Index task. Null when not measured. type: - number - "null" reasoning_cost: description: Weighted average reasoning-token cost in USD per Intelligence Index task. Null when not measured. type: - number - "null" answer_cost: description: Weighted average answer-token cost in USD per Intelligence Index task. Null when not measured. type: - number - "null" required: - total_cost - input_cost - reasoning_cost - answer_cost IntelligenceIndexTokenCounts: type: object properties: input_tokens: type: integer answer_tokens: type: integer output_tokens: type: integer reasoning_tokens: type: integer required: - input_tokens - answer_tokens - output_tokens - reasoning_tokens OmniscienceDomainBreakdown: type: object properties: accuracy: type: number omniscience: type: number hallucination_rate: type: number required: - accuracy - omniscience - hallucination_rate OmniscienceBreakdown: type: object properties: total: $ref: "#/components/schemas/OmniscienceDomainBreakdown" law: oneOf: - $ref: "#/components/schemas/OmniscienceDomainBreakdown" - type: "null" health: oneOf: - $ref: "#/components/schemas/OmniscienceDomainBreakdown" - type: "null" business: oneOf: - $ref: "#/components/schemas/OmniscienceDomainBreakdown" - type: "null" humanities_and_social_sciences: oneOf: - $ref: "#/components/schemas/OmniscienceDomainBreakdown" - type: "null" science_engineering_and_mathematics: oneOf: - $ref: "#/components/schemas/OmniscienceDomainBreakdown" - type: "null" software_engineering: oneOf: - $ref: "#/components/schemas/OmniscienceDomainBreakdown" - type: "null" required: - total OpennessBreakdown: type: object properties: weights_access: type: - number - "null" weights_license: type: - number - "null" data_pretrain_access: type: - number - "null" data_pretrain_license: type: - number - "null" data_posttrain_access: type: - number - "null" data_posttrain_license: type: - number - "null" methodology_license: type: - number - "null" methodology_disclosure: type: - number - "null" artificial_analysis_openness_index: type: - number - "null" ModelProviderDataExternal: type: object properties: id: type: string name: type: string slug: type: string endpoint_label: description: Label that distinguishes this endpoint from other endpoints of the same provider and model, e.g. "(FAST)" or "(FP8)". A provider that serves a model through several endpoints has one entry per endpoint. Null when the endpoint has no label. type: - string - "null" pricing: $ref: "#/components/schemas/Pricing" performance: $ref: "#/components/schemas/ProviderModelPerformance" performance_v3: oneOf: - $ref: "#/components/schemas/ProviderModelPerformanceV3" - type: "null" description: Performance v3 block for this endpoint. Null when the endpoint has no measurements in the latest window. context_window_tokens: description: Null when not set for the (provider, model) pair. type: - integer - "null" required: - id - name - slug - endpoint_label - pricing - performance - performance_v3 - context_window_tokens ProviderModelPerformance: type: object properties: percentile_05_output_tokens_per_second: type: - number - "null" quartile_25_output_tokens_per_second: type: - number - "null" median_output_tokens_per_second: type: - number - "null" quartile_75_output_tokens_per_second: type: - number - "null" percentile_95_output_tokens_per_second: type: - number - "null" percentile_05_time_to_first_token_seconds: type: - number - "null" quartile_25_time_to_first_token_seconds: type: - number - "null" median_time_to_first_token_seconds: type: - number - "null" quartile_75_time_to_first_token_seconds: type: - number - "null" percentile_95_time_to_first_token_seconds: type: - number - "null" median_time_to_first_answer_token_seconds: type: - number - "null" median_end_to_end_response_time_seconds: type: - number - "null" required: - percentile_05_output_tokens_per_second - quartile_25_output_tokens_per_second - median_output_tokens_per_second - quartile_75_output_tokens_per_second - percentile_95_output_tokens_per_second - percentile_05_time_to_first_token_seconds - quartile_25_time_to_first_token_seconds - median_time_to_first_token_seconds - quartile_75_time_to_first_token_seconds - percentile_95_time_to_first_token_seconds - median_time_to_first_answer_token_seconds - median_end_to_end_response_time_seconds description: Every field returns `null` from December 10, 2026. Use `performance_v3` ([methodology](https://artificialanalysis.ai/methodology/performance-benchmarking)). Performance metrics for a single (provider, model) pair. Individual fields are null when no measurements are available for that metric. deprecated: true LLMProvidersResponse: type: object properties: tier: type: string enum: - commercial pagination: $ref: "#/components/schemas/PaginationMeta" data: type: array items: $ref: "#/components/schemas/ProviderDataExternalList" required: - tier - pagination - data LLMProviderItemResponse: type: object properties: tier: type: string enum: - commercial data: $ref: "#/components/schemas/ProviderDataExternalDetail" required: - tier - data ProviderDataExternalList: type: object properties: id: type: string name: type: string slug: type: string country: description: ISO 3166-1 alpha-2 country code of the provider. Null when unknown. type: - string - "null" models: type: array items: $ref: "#/components/schemas/ProviderModelExternalList" required: - id - name - slug - models ProviderDataExternalDetail: type: object properties: id: type: string name: type: string slug: type: string country: description: ISO 3166-1 alpha-2 country code of the provider. Null when unknown. type: - string - "null" models: type: array items: $ref: "#/components/schemas/ProviderModelExternalDetail" required: - id - name - slug - models ProviderModelExternalList: type: object properties: id: type: string name: type: string slug: type: string endpoint_label: description: Label that distinguishes this endpoint from other endpoints of the same provider and model, e.g. "(FAST)" or "(FP8)". A provider that serves a model through several endpoints has one entry per endpoint. Null when the endpoint has no label. type: - string - "null" pricing: $ref: "#/components/schemas/Pricing" context_window_tokens: description: Null when not set for the (provider, model) pair. type: - integer - "null" performance: oneOf: - $ref: "#/components/schemas/ProviderModelPerformanceList" - type: "null" performance_v3: oneOf: - $ref: "#/components/schemas/ProviderModelPerformanceV3List" - type: "null" description: Performance v3 headline means, endpoint error rate and window date for this endpoint. The provider detail endpoint (`/api/v2/language/providers/{slug}`) adds per-workload metrics; see `ProviderModelPerformanceV3`. Null when the endpoint has no measurements in the latest window. required: - id - name - slug - endpoint_label - pricing - context_window_tokens - performance - performance_v3 ProviderModelExternalDetail: type: object properties: id: type: string name: type: string slug: type: string endpoint_label: description: Label that distinguishes this endpoint from other endpoints of the same provider and model, e.g. "(FAST)" or "(FP8)". A provider that serves a model through several endpoints has one entry per endpoint. Null when the endpoint has no label. type: - string - "null" pricing: $ref: "#/components/schemas/Pricing" context_window_tokens: description: Null when not set for the (provider, model) pair. type: - integer - "null" performance: oneOf: - $ref: "#/components/schemas/ProviderModelPerformanceDetail" - type: "null" performance_v3: oneOf: - $ref: "#/components/schemas/ProviderModelPerformanceV3" - type: "null" description: Full Performance v3 block for this endpoint, including per-workload metrics. The providers list returns the headline only; see `ProviderModelPerformanceV3List`. Null when the endpoint has no measurements in the latest window. required: - id - name - slug - endpoint_label - pricing - context_window_tokens - performance - performance_v3 ProviderModelPerformanceList: type: object properties: median_output_tokens_per_second: type: - number - "null" median_time_to_first_token_seconds: type: - number - "null" median_time_to_first_answer_token_seconds: type: - number - "null" median_end_to_end_response_time_seconds: type: - number - "null" required: - median_output_tokens_per_second - median_time_to_first_token_seconds - median_time_to_first_answer_token_seconds - median_end_to_end_response_time_seconds description: Every field returns `null` from December 10, 2026. Use `performance_v3` ([methodology](https://artificialanalysis.ai/methodology/performance-benchmarking)). Median performance metrics returned on the providers list endpoint. Individual fields are null when no measurements are available for that metric. deprecated: true ProviderModelPerformanceDetail: type: object properties: median_output_tokens_per_second: type: - number - "null" median_time_to_first_token_seconds: type: - number - "null" percentile_05_output_tokens_per_second: type: - number - "null" percentile_95_output_tokens_per_second: type: - number - "null" quartile_25_output_tokens_per_second: type: - number - "null" quartile_75_output_tokens_per_second: type: - number - "null" percentile_05_time_to_first_token_seconds: type: - number - "null" percentile_95_time_to_first_token_seconds: type: - number - "null" quartile_25_time_to_first_token_seconds: type: - number - "null" quartile_75_time_to_first_token_seconds: type: - number - "null" median_time_to_first_answer_token_seconds: type: - number - "null" median_end_to_end_response_time_seconds: type: - number - "null" required: - median_output_tokens_per_second - median_time_to_first_token_seconds - percentile_05_output_tokens_per_second - percentile_95_output_tokens_per_second - quartile_25_output_tokens_per_second - quartile_75_output_tokens_per_second - percentile_05_time_to_first_token_seconds - percentile_95_time_to_first_token_seconds - quartile_25_time_to_first_token_seconds - quartile_75_time_to_first_token_seconds - median_time_to_first_answer_token_seconds - median_end_to_end_response_time_seconds description: Every field returns `null` from December 10, 2026. Use `performance_v3` ([methodology](https://artificialanalysis.ai/methodology/performance-benchmarking)). Full performance metrics returned on the provider detail endpoint. Includes everything from the list endpoint plus percentile/quartile breakdowns. Individual fields are null when no measurements are available for that metric. deprecated: true PerformanceSnapshot: type: object properties: date: type: string description: Start date of this data point's period (YYYY-MM-DD). median_output_tokens_per_second: type: - number - "null" median_time_to_first_token_seconds: type: - number - "null" median_time_to_first_answer_token_seconds: type: - number - "null" median_end_to_end_response_time_seconds: type: - number - "null" required: - date - median_output_tokens_per_second - median_time_to_first_token_seconds - median_time_to_first_answer_token_seconds - median_end_to_end_response_time_seconds description: Every metric returns `null` from December 10, 2026. Use `series_v3`. A single performance data point for a given date or interval period. deprecated: true PerformanceV3Snapshot: type: object properties: date: type: string description: First day of the period (YYYY-MM-DD). A daily point covers the trailing 5-day window ending that day, as the website does. A weekly or monthly point covers the 7 or 30 days from that day. mean_output_tokens_per_second: description: Headline end-to-end output speed. Mean over the workloads measured. type: - number - "null" mean_time_to_first_token_seconds: description: Mean over the workloads measured. type: - number - "null" mean_time_to_first_answer_token_seconds: description: Mean over the workloads measured. type: - number - "null" endpoint_error_rate: description: Endpoint error samples over all samples across the workloads measured (0-1). Null when no workload measured an output speed. type: - number - "null" required: - date - mean_output_tokens_per_second - mean_time_to_first_token_seconds - mean_time_to_first_answer_token_seconds - endpoint_error_rate description: "A Performance v3 data point for one endpoint: end-to-end speed averaged over the single-request and multi-turn agentic workloads measured in the period. Data starts on September 15, 2026." ModelPerformanceOverTimeResponse: type: object properties: tier: type: string enum: - commercial data: type: object properties: id: type: string slug: type: string name: type: string providers: type: array items: type: object properties: id: type: string slug: type: string name: type: string endpoint_label: description: Label that distinguishes this endpoint from other endpoints of the same provider and model, e.g. "(FAST)" or "(FP8)". A provider that serves a model through several endpoints has one entry per endpoint. Null when the endpoint has no label. type: - string - "null" series: type: array items: $ref: "#/components/schemas/PerformanceSnapshot" deprecated: true series_v3: type: array items: $ref: "#/components/schemas/PerformanceV3Snapshot" required: - id - slug - name - endpoint_label - series - series_v3 required: - id - slug - name - providers required: - tier - data ProviderPerformanceOverTimeResponse: type: object properties: tier: type: string enum: - commercial data: type: object properties: id: type: string name: type: string slug: type: string models: type: array items: type: object properties: id: type: string slug: type: string name: type: string endpoint_label: description: Label that distinguishes this endpoint from other endpoints of the same provider and model, e.g. "(FAST)" or "(FP8)". A provider that serves a model through several endpoints has one entry per endpoint. Null when the endpoint has no label. type: - string - "null" series: type: array items: $ref: "#/components/schemas/PerformanceSnapshot" deprecated: true series_v3: type: array items: $ref: "#/components/schemas/PerformanceV3Snapshot" required: - id - slug - name - endpoint_label - series - series_v3 required: - id - name - slug - models required: - tier - data MeasurementsResponse: type: object properties: tier: type: string enum: - commercial data: type: object properties: id: type: string name: type: string slug: type: string models: type: array items: $ref: "#/components/schemas/MeasurementModel" required: - id - name - slug - models required: - tier - data MeasurementModel: type: object properties: id: type: string name: type: string slug: type: string measurements: type: array items: $ref: "#/components/schemas/Measurement" deprecated: true measurements_v3: type: array items: $ref: "#/components/schemas/MeasurementV3" required: - id - name - slug - measurements - measurements_v3 Measurement: type: object properties: id: type: string date: type: string description: Timestamp when the measurement was recorded. success: type: boolean output_tokens_per_second: type: - number - "null" time_to_first_token_seconds: type: - number - "null" time_to_first_answer_token: type: - number - "null" end_to_end_response_time_seconds: type: - number - "null" required: - id - date - success - output_tokens_per_second - time_to_first_token_seconds - time_to_first_answer_token - end_to_end_response_time_seconds description: Every metric returns `null` from December 10, 2026. Use `measurements_v3`. deprecated: true MeasurementV3: type: object properties: run_id: type: string date: type: string description: Timestamp when the run was measured. workload: type: string enum: - 1k - 10k - 100k - multi_turn description: "Workload measured: `1k`, `10k` or `100k` input tokens, or `multi_turn`, one of the multi-turn agentic workloads." endpoint_label: description: Label that distinguishes this endpoint from other endpoints of the same provider and model, e.g. "(FAST)" or "(FP8)". A provider that serves a model through several endpoints has one entry per endpoint. Null when the endpoint has no label. type: - string - "null" median_output_tokens_per_second: description: End-to-end output speed, median over the run's samples. type: - number - "null" median_time_to_first_token_seconds: type: - number - "null" median_time_to_first_answer_token_seconds: type: - number - "null" median_end_to_end_time_seconds: type: - number - "null" samples: type: number description: Samples in the run. endpoint_error_samples: type: number description: Samples that failed with an endpoint error. required: - run_id - date - workload - endpoint_label - median_output_tokens_per_second - median_time_to_first_token_seconds - median_time_to_first_answer_token_seconds - median_end_to_end_time_seconds - samples - endpoint_error_samples description: One Performance v3 run of one workload on one endpoint, as medians over its samples. ArenaModelCreator: type: object properties: id: type: string name: type: string required: - id - name ImageArenaResponse: type: object properties: tier: type: string enum: - pro - commercial data: type: array items: $ref: "#/components/schemas/ImageArenaResult" required: - tier - data ImageArenaFreeResponse: type: object properties: tier: type: string enum: - free - pro - commercial data: type: array items: $ref: "#/components/schemas/ImageArenaResultFree" required: - tier - data ImageArenaResult: type: object properties: id: type: string name: type: string slug: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" elo: type: number description: Arena Elo rating. ci_95: description: 95% confidence interval. Null if unavailable. type: - number - "null" rank: description: Rank among current models by Elo. type: - integer - "null" samples: type: integer description: Total number of arena samples. release_date: description: Release date. type: - string - "null" price_per_1k_images: description: Price per 1,000 images. type: - number - "null" open_weights_url: description: URL to open weights; can be used to infer if open weights. type: - string - "null" categories: type: array items: $ref: "#/components/schemas/ArenaCategory" description: Per-category Elo breakdowns. Only present when include_categories=true. required: - id - name - slug - model_creator - elo - ci_95 - rank - samples - release_date - price_per_1k_images - open_weights_url ImageArenaResultFree: type: object properties: id: type: string name: type: string slug: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" elo: type: number description: Arena Elo rating. ci_95: description: 95% confidence interval. Null if unavailable. type: - number - "null" required: - id - name - slug - model_creator - elo - ci_95 VideoArenaResponse: type: object properties: tier: type: string enum: - pro - commercial data: type: array items: $ref: "#/components/schemas/VideoArenaResultPaid" required: - tier - data VideoArenaFreeResponse: type: object properties: tier: type: string enum: - free - pro - commercial data: type: array items: $ref: "#/components/schemas/VideoArenaResultFree" required: - tier - data VideoArenaResultFree: type: object properties: id: type: string name: type: string slug: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" elo: type: number description: Arena Elo rating. ci_95: description: 95% confidence interval. Null if unavailable. type: - number - "null" required: - id - name - slug - model_creator - elo - ci_95 VideoArenaResultPaid: type: object properties: id: type: string name: type: string slug: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" elo: type: number description: Arena Elo rating. ci_95: description: 95% confidence interval. Null if unavailable. type: - number - "null" rank: description: Rank among current models by Elo. type: - integer - "null" samples: type: integer description: Total number of arena samples. release_date: description: Release date. type: - string - "null" price_per_minute: description: Price per minute of video. type: - number - "null" open_weights_url: description: URL to open weights; can be used to infer if open weights. type: - string - "null" categories: type: array items: $ref: "#/components/schemas/ArenaCategory" description: Per-category Elo breakdowns. Only present when include_categories=true. required: - id - name - slug - model_creator - elo - ci_95 - rank - samples - release_date - price_per_minute - open_weights_url ArenaCategory: type: object properties: label: type: string description: Category label. group: type: string description: Taxonomy the category belongs to, e.g. "use_case", "capability", "style" or "editing_action". Use it to read one breakdown; labels from different groups can sit side by side. elo: type: number ci_95: type: - number - "null" samples: type: integer required: - label - group - elo - ci_95 - samples TextToSpeechResponse: type: object properties: tier: type: string enum: - pro - commercial data: type: array items: $ref: "#/components/schemas/TextToSpeechResult" required: - tier - data TextToSpeechFreeResponse: type: object properties: tier: type: string enum: - free - pro - commercial data: type: array items: $ref: "#/components/schemas/TextToSpeechResultFree" required: - tier - data TextToSpeechResult: type: object properties: id: type: string name: type: string slug: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" elo: type: number description: Arena Elo rating. ci_95: description: 95% confidence interval. Null if unavailable. type: - number - "null" samples: type: integer description: Total number of arena samples. release_date: description: Release date. type: - string - "null" price_per_1m_characters: description: Price per 1 million characters. type: - number - "null" required: - id - name - slug - model_creator - elo - ci_95 - samples - release_date - price_per_1m_characters TextToSpeechResultFree: type: object properties: id: type: string name: type: string slug: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" elo: type: number description: Arena Elo rating. ci_95: description: 95% confidence interval. Null if unavailable. type: - number - "null" required: - id - name - slug - model_creator - elo - ci_95 SpeechToSpeechResponse: type: object properties: tier: type: string enum: - pro - commercial data: type: array items: $ref: "#/components/schemas/SpeechToSpeechResult" required: - tier - data SpeechToSpeechFreeResponse: type: object properties: tier: type: string enum: - free - pro - commercial data: type: array items: $ref: "#/components/schemas/SpeechToSpeechResultFree" required: - tier - data SpeechToSpeechResult: type: object properties: id: type: string name: type: string slug: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" bba_score: description: Big Bench Audio score. type: - number - "null" fdb_score: description: Full Duplex Bench score. type: - number - "null" tau_voice_score: description: 𝜏-Voice score (the model's best result across providers). Null when the model has no 𝜏-Voice measurements. type: - number - "null" providers: type: array items: $ref: "#/components/schemas/SpeechToSpeechProvider" required: - id - name - slug - model_creator - bba_score - fdb_score - tau_voice_score - providers SpeechToSpeechResultFree: type: object properties: id: type: string name: type: string slug: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" bba_score: description: Big Bench Audio score. type: - number - "null" fdb_score: description: Full Duplex Bench score. type: - number - "null" tau_voice_score: description: 𝜏-Voice score (the model's best result across providers). Null when the model has no 𝜏-Voice measurements. type: - number - "null" required: - id - name - slug - model_creator - bba_score - fdb_score - tau_voice_score SpeechToSpeechProvider: type: object properties: id: type: string name: type: string slug: type: string price_per_hour_input: description: Price per hour of input audio. type: - number - "null" price_per_hour_output: description: Price per hour of output audio. type: - number - "null" time_to_first_audio_seconds: description: Time to first audio output in seconds. type: - number - "null" required: - id - name - slug - price_per_hour_input - price_per_hour_output - time_to_first_audio_seconds SpeechToTextModelsResponse: type: object properties: tier: type: string enum: - pro - commercial data: type: array items: $ref: "#/components/schemas/SpeechToTextModelResult" required: - tier - data SpeechToTextModelsFreeResponse: type: object properties: tier: type: string enum: - free - pro - commercial data: type: array items: $ref: "#/components/schemas/SpeechToTextModelResultFree" required: - tier - data SpeechToTextModelResult: type: object properties: id: type: string name: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" aa_wer_index: description: AA Word Error Rate index. type: - number - "null" aa_agenttalk: description: AA AgentTalk WER. type: - number - "null" voxpopuli_cleaned_aa: description: VoxPopuli cleaned WER. type: - number - "null" earnings_22_cleaned_aa: description: Earnings22 cleaned WER. type: - number - "null" open_weights: description: Whether the model has open weights. Null when unknown. type: - boolean - "null" providers: type: array items: $ref: "#/components/schemas/SpeechToTextProvider" description: Per-provider data. required: - id - name - model_creator - aa_wer_index - aa_agenttalk - voxpopuli_cleaned_aa - earnings_22_cleaned_aa - open_weights - providers SpeechToTextModelResultFree: type: object properties: id: type: string name: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" aa_wer_index: description: AA Word Error Rate index. type: - number - "null" required: - id - name - model_creator - aa_wer_index SpeechToTextProvider: type: object properties: id: type: string name: type: string slug: type: string price_per_1k_minutes: type: - number - "null" median_speed_factor: type: - number - "null" aa_wer_index: type: - number - "null" aa_agenttalk: type: - number - "null" voxpopuli_cleaned_aa: type: - number - "null" earnings_22_cleaned_aa: type: - number - "null" required: - id - name - slug - price_per_1k_minutes - median_speed_factor - aa_wer_index - aa_agenttalk - voxpopuli_cleaned_aa - earnings_22_cleaned_aa MusicArenaResponse: type: object properties: tier: type: string enum: - pro - commercial data: type: array items: $ref: "#/components/schemas/MusicArenaResult" required: - tier - data MusicArenaFreeResponse: type: object properties: tier: type: string enum: - free - pro - commercial data: type: array items: $ref: "#/components/schemas/MusicArenaResultFree" required: - tier - data MusicArenaResult: type: object properties: id: type: string name: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" elo: type: number description: Arena Elo rating. ci_95: description: 95% confidence interval. Null if unavailable. type: - number - "null" samples: type: integer description: Total number of arena samples. genres: type: array items: $ref: "#/components/schemas/MusicGenre" description: Per-genre Elo breakdowns. Only present when include_genres=true. required: - id - name - model_creator - elo - ci_95 - samples MusicArenaResultFree: type: object properties: id: type: string name: type: string model_creator: $ref: "#/components/schemas/ArenaModelCreator" elo: type: number description: Arena Elo rating. ci_95: description: 95% confidence interval. Null if unavailable. type: - number - "null" required: - id - name - model_creator - elo - ci_95 MusicGenre: type: object properties: label: type: string description: Genre name (e.g. "Classical", "Electronic", "Pop"). elo: type: number ci_95: type: - number - "null" samples: type: integer required: - label - elo - ci_95 - samples