AA-3D: Interactive 3D Experience Benchmark
AA-3D
AA-3D Leaderboard
Range | Model | Elo | |
|---|---|---|---|
| 1 | 1-2 |  Claude Sonnet 5.5 (Max, Default Fallback) | 1199±3395% interval 1166 to 1231 |
| 2 | 1-3 | Claude Opus 5.5 (Max, Default Fallback) | 1196±3395% interval 1164 to 1229 |
| 3 | 3-5 |  GPT-6 Astra (Max) | 1164±3295% interval 1132 to 1196 |
| 4 | 3-6 |  Gemini 4 Argon (High) | 1153±3395% interval 1120 to 1186 |
| 5 | 3-8 | Claude Fable 5.1 (Max, Default Fallback) | 1135±3295% interval 1103 to 1167 |
| 6 | 4-8 | GPT-6 Sol (Max) | 1121±3295% interval 1089 to 1153 |
| 7 | 5-8 | GPT-6.1 Sol (Max) | 1119±3295% interval 1088 to 1151 |
| 8 | 5-8 |  Kimi K3.1 Pro | 1110±3395% interval 1078 to 1143 |
| 9 | 9 |  Muse Spark 1.3 (Max) | 1033±3195% interval 1003 to 1064 |
| 10 | 10 | GPT-6 Luna (Max) | 1000AnchorFixed at 1000 as the reference point for this board's Elo scale. |
| 11 | 10-14 | Claude Haiku 5.5 (Max) | 998±3195% interval 967 to 1028 |
| 12 | 10-15 |  Grok 4.7 (Xhigh) | 995±3195% interval 964 to 1025 |
| 13 | 10-15 |  MiMo-V2.6-Pro | 983±3195% interval 952 to 1014 |
| 14 | 10-15 | Kimi K3 (Max) | 982±3195% interval 951 to 1013 |
| 15 | 12-17 | Gemini 3.8 Flash (High) | 966±3195% interval 935 to 997 |
| 16 | 15-17 |  Qwen3.8 Max (0902) | 945±3195% interval 914 to 976 |
| 17 | 15-17 |  GLM-5.3-Flash | 937±3195% interval 906 to 967 |
| 18 | 18-19 |  Mistral Large 4 Preview | 870±3195% interval 839 to 901 |
| 19 | 18-19 |  Step 5 Preview | 854±3295% interval 821 to 886 |
| 20 | 20 | GLM-5.3 (Max) | 783±3495% interval 749 to 817 |
| 21 | 21 |  MiniMax-M3 | 740±3395% interval 707 to 773 |
| 22 | 22-23 |  Inkling (Xhigh) | 583±4095% interval 543 to 623 |
| 23 | 22-23 |  Nemotron 3 Ultra 550B A55B (Reasoning) | 573±4395% interval 529 to 616 |
AA-3D Elo
AA-3D Elo
AA-3D: Elo vs. Cost per Task
Pick a domain, a task and a model to load the scene that model built.
Task
Model
Create an interactive adult man performing weightlifting exercises, with views that expose his muscles, skeleton and organs in anatomically accurate form.
23 of 23 models have a scene for this task in blind voting.
Elo vs. Intelligence Index
AA-3D Elo vs. Intelligence Index
Win Rate by Domain
Each cell shows the share of votes a model won in one task domain, from the votes behind the current leaderboard. The opponent-balanced view takes the mean of the model's win rates against each opponent in the domain, so every opponent counts equally.
AA-3D Win Rate by Domain
Cost
AA-3D: Cost per Task
Token Usage
AA-3D: Output Tokens per Task
Turns per Scene
AA-3D Turns per Scene
Tool Use
AA-3D Tool Calls per Scene
How AA-3D Works
Each model works as an agent in an offline sandbox. It gets a written brief for one interactive scene, plus a delivery contract that every task shares. It writes code, renders the scene, looks at its own screenshots and revises. The sandbox has Blender, Rapier physics, Python, a headless browser and a library of CC0 assets. A run stops after three hours or 250 turns, and the result is an index.html page that runs the scene in a browser.
Inside an Agent Run
These replays use the AA-3D launch data. Pick a task to replay every model's run on a shared clock. Hover over a lane to see each action, and click it to see where the time went.
Watch a Scene Being Built
Score vs. Release Date
AA-3D: Score vs. Release Date
Frequently Asked Questions
AA-3D v1.0 measures how well AI models build interactive 3D experiences. Each model receives a written brief and builds a single web page with three.js in an offline sandbox. The benchmark has 55 tasks across 7 domains. Read the AA-3D methodology
Models are ranked using an Elo rating system derived from blind comparisons judged by our recruited human panel. Panelists interact with two scenes built for the same task and choose the one they prefer, without knowing which model built which. Higher Elo scores indicate a model is preferred more often. Try AA-3D Voting
Explore Evaluations
Artificial Analysis Intelligence Index v5.0A composite benchmark aggregating nine challenging evaluations to provide a holistic measure of AI capabilities across knowledge work, coding, science, and general knowledge.
AA-Briefcase v1.1: Agentic Knowledge Work BenchmarkA private evaluation developed by Artificial Analysis for frontier agentic capability in long-horizon knowledge work, testing agents on realistic business workflows that require deliverables such as spreadsheets, presentations, and memos.
AA-3D: Interactive 3D Experience BenchmarkA benchmark developed by Artificial Analysis for how well AI models build interactive 3D experiences. Models work as agents in an offline sandbox via Stirrup to turn a written brief into a three.js scene, with Elo ratings derived from blind pairwise votes.
APEX-Agents-AA BenchmarkArtificial Analysis' implementation of the APEX-Agents benchmark, testing AI agents on long-horizon, cross-application tasks in professional-services environments with realistic application tooling.
AA-AnalystAgent BenchmarkArtificial Analysis' data analysis benchmark, testing AI agents on their ability to work with spreadsheets and documents to answer quantitative questions a Business Analyst or Data Analyst would face day-to-day.
AutomationBench-AA: Agentic SaaS Workflow BenchmarkA benchmark measuring agentic task completion across simulated SaaS application environments, scoring the share of each task's objectives completed without guardrail violations.
Harvey LAB-AA v1.1 BenchmarkArtificial Analysis' implementation of Harvey's Legal Agent Benchmark (LAB), testing AI agents on real-world legal work from Harvey's dataset of 120 private tasks spanning 24 legal practice areas. v1.1, built in collaboration with Harvey, checks every deliverable against the task's source documents for hallucinations, so work that passes every rubric criterion but contains a material hallucination receives no credit. A panel of three judges grades each deliverable criterion-by-criterion.
GDP.pdf BenchmarkArtificial Analysis' implementation of Surge AI's GDP.pdf benchmark, testing whether language models can reason over long, real-world professional documents and satisfy detailed task-specific criteria.
EnterpriseOps-Gym-AA BenchmarkArtificial Analysis' independent implementation of ServiceNow's EnterpriseOps-Gym, an agentic benchmark testing whether LLM agents can complete stateful, multi-step enterprise workflows across eight business domains via live tool use, graded on the final state of the underlying databases.
Terminal-Bench 4.0 BenchmarkA harder 66-task benchmark of complex terminal work across software, machine learning, science, operations, security, hardware, and media, with recalibrated compute and time allowances and improved instructions, environments, and verifiers.
Terminal-Bench-Science 0.1 BenchmarkA 70-task benchmark of research workflows authored and reviewed by domain experts across the life, physical, earth, mathematical, and engineering sciences, each completed in a terminal and checked by its own set of tests.
FrontierCode 2.0 BenchmarkArtificial Analysis' implementation of Cognition's FrontierCode 2.0, a private agentic coding benchmark of 132 tasks in real open-source repositories, graded on whether a maintainer would merge the agent's change.
HLE-Diamond BenchmarkA refined and updated 1,000-question set of Humanity's Last Exam, split evenly between reasoning and knowledge questions, with 272 questions that include an image.
AA-PersonalityBench: AI Personality & Communication Style BenchmarkMeasures how models communicate rather than whether they are correct: 30 judged personality traits rolled up into 8 indexes, plus 21 measured communication characteristics.
AA-Omniscience v1.1: Knowledge and Hallucination BenchmarkA benchmark measuring factual recall and hallucination across various economically relevant domains.
The most challenging 198 questions from GPQA, where PhD experts achieve 65% accuracy but skilled non-experts only reach 34% despite web access.
ITBench-AA BenchmarkArtificial Analysis' implementation of IBM's ITBench benchmark, testing AI agents on Kubernetes incident root-cause analysis from offline incident snapshots. The agent inspects alerts, events, traces, and topology and identifies the contributing-factor entities (deployments, pods, namespaces, network policies, etc.) responsible for the failure.
MMMU-Pro BenchmarkAn enhanced MMMU benchmark that eliminates shortcuts and guessing strategies to more rigorously test multimodal models across 30 academic disciplines.
IFBench BenchmarkA benchmark evaluating precise instruction-following generalization on 58 diverse, verifiable out-of-domain constraints that test models' ability to follow specific output requirements.
MLCR-AA: Medical Long Context Reasoning BenchmarkAn open benchmark from Wisedocs measuring how well models reason over long, fragmented medical records, performing the multi-document synthesis claims professionals rely on when reviewing insurance and healthcare cases.
Code2Video-AA BenchmarkArtificial Analysis' implementation of HeyGen's Code2Video Bench, evaluating how well models create short product and brand videos from 114 design briefs using HyperFrames HTML.
HealthBench BenchmarkOpenAI's HealthBench: 5,000 health conversations graded against physician-written rubrics, from emergency referrals to global health.
CritPt v2 BenchmarkCritPt v2 is the revised version of CritPt, a benchmark designed to test LLMs on research-level physics reasoning tasks, featuring 70 composite research challenges. Models solve problems with a code-execution tool in an offline Linux sandbox.
GDPval-AA v2.1 BenchmarkGDPval-AA v2.1 is Artificial Analysis' evaluation framework for OpenAI's GDPval dataset. It tests AI models on real-world tasks across 44 occupations and 9 major industries. Models are given shell access and web browsing capabilities in an agentic loop via Stirrup to solve tasks, with Elo ratings derived from blind pairwise comparisons.
𝜏³-Banking BenchmarkA fintech customer-support benchmark from the 𝜏-Knowledge framework that tests whether agents can navigate a large unstructured knowledge base and execute multi-step tool calls to resolve realistic banking workflows.
Terminal-Bench 2.1 BenchmarkA verified refresh of Terminal-Bench 2.0: 89 curated tasks across software engineering, system administration, data processing, model training, and security, with environment and instruction fixes so scores reflect agent capability rather than environment gaps.
FrontierCode 1.1 BenchmarkArtificial Analysis' implementation of Cognition's FrontierCode 1.1, a private agentic coding benchmark that tests models on completing coding tasks in a way that would actually get merged by maintainers of real-world repositories.
AA-LCR v1.1: Long Context Reasoning BenchmarkA challenging benchmark measuring language models' ability to extract, reason about, and synthesize information from long-form documents ranging from 10k to 100k tokens (measured using the cl100k_base tokenizer).
SciCode BenchmarkA scientist-curated coding benchmark featuring 288 test set subproblems from 80 laboratory problems across 16 scientific disciplines.
Humanity's Last Exam BenchmarkA frontier-level benchmark with 2,500 expert-vetted questions across mathematics, sciences, and humanities, designed to be the final closed-ended academic evaluation.
CritPt BenchmarkA benchmark designed to test LLMs on research-level physics reasoning tasks, featuring 71 composite research challenges.
Terminal-Bench Hard BenchmarkAn agentic benchmark evaluating AI capabilities in terminal environments through software engineering, system administration, and data processing tasks.
𝜏²-Bench Telecom BenchmarkA dual-control conversational AI benchmark simulating technical support scenarios where both agent and user must coordinate actions to resolve telecom service issues.
MMLU-Pro BenchmarkAn enhanced version of MMLU with 12,000 graduate-level questions across 14 subject areas, featuring ten answer options and deeper reasoning requirements.
LiveCodeBench BenchmarkA contamination-free coding benchmark that continuously harvests fresh competitive programming problems from LeetCode, AtCoder, and CodeForces, evaluating code generation, self-repair, and execution.
MATH-500 BenchmarkA 500-problem subset from the MATH dataset, featuring competition-level mathematics across six domains including algebra, geometry, and number theory.
AIME 2025 BenchmarkAll 30 problems from the 2025 American Invitational Mathematics Examination, testing olympiad-level mathematical reasoning with integer answers from 000-999.
Global-MMLU-Lite BenchmarkA lightweight, multilingual version of MMLU, designed to evaluate knowledge and reasoning skills across a diverse range of languages and cultural contexts.