DeepInfra: Models Intelligence, Performance & Price
Analysis of DeepInfra's models across key metrics including quality, price, output speed, latency, context window & more. This analysis is intended to support you in choosing the best model provided by DeepInfra for your use-case.
Most Intelligent
Intelligence index
Total 100 models
Fastest
Output speed
Total 100 models
Lowest Price
Blended price (per 1M tokens)
Total 100 models
DeepInfra offers 100 models, each with different intelligence, performance, and pricing characteristics. Below is a comparison of the key metrics across models.
- For intelligence, the top models on DeepInfra are GLM-5.2 (max) (FP4) (53), DeepSeek V4 Flash 0731 (max) (52), and MiniMax-M3 (45).
- For output speed, the fastest models are Nemotron 3.5 Lightning (NVFP4) (666 t/s), Ling 3.0 Flash (289 t/s), and Step 3.7 Flash (176 t/s). Speed varies significantly across models, with a 337% difference between the fastest and slowest.
- For latency, Nemotron 3 Nano (0.40s), Qwen3.5 35B A3B FP8 (0.51s), and Qwen3 30B (FP8) (0.52s) offer the lowest time to first answer token.
- For pricing, Llama 3.1 8B (Turbo, FP8) ($0.02), Llama 3.1 8B ($0.02), and Gemma 4 E4B ($0.03) offer the lowest blended prices per 1M tokens.
- For context window size, GLM-5.2 (max) (FP4) (1M), DeepSeek V4 Pro (max) (FP4) (1M), and DeepSeek V4 Flash (max) (FP4) (1M) support the largest context windows on DeepInfra.
Highlights
Intelligence Evaluations
Artificial Analysis Intelligence Index
Intelligence Evaluations
Agentic real-world work tasks, (Elo-500)/2000
Agentic tool use
Agentic coding & terminal use
Coding
Reasoning & knowledge
Scientific reasoning
Physics reasoning
Knowledge
1 - hallucination rate
Long context reasoning
Agentic knowledge work, Elo
Agentic SaaS workflows
Legal agentic work, criterion pass rate
Agentic business operations
Quantitative analysis on spreadsheets & documents
Instruction following
Long-horizon agentic tasks
Kubernetes incident root-cause analysis
Visual reasoning
Intelligence Index vs. Price
Context Window
Context Window
Pricing
Intelligence Index vs. Price
Performance Summary
Output Speed vs. Price
Speed
Measured by Output Speed (tokens per second)
Output Speed
Latency
Measured by Time (seconds) to First Token
Latency: Time To First Answer Token
End-to-End Response Time
Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed
End-to-End Response Time vs. Price
Further Analysis | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
GLM-5.2 (max) (FP4) | 1.05M | Open | 53 | $0.25 | 42 | 1.50 | 61.53 | 48.02 | |||
DeepSeek V4 Flash 0731 (max) | 1M | Open | 52 | $0.06 | 60 | 0.89 | 42.30 | 33.13 | |||
MiniMax-M3 | 524k | Open | 45 | $0.13 | 51 | 1.64 | 50.48 | 39.07 | |||
DeepSeek V4 Pro (max) (FP4) | 1.05M | Open | 45 | $0.27 | 37 | 1.66 | 135.11 | 119.77 | |||
Kimi K2.6 (FP4) | 262k | Open | 45 | $0.27 | 19 | 1.92 | 265.74 | 237.18 | |||
DeepSeek V4 Pro (high) (FP4) | 65.5k | Open | 44 | $0.26 | 37 | 1.41 | 68.94 | 53.98 | |||
Kimi K2.7 Code | 262k | Open | 43 | $0.18 | 84 | 0.65 | 33.01 | 26.43 | |||
 MiMo-V2.5-Pro | 65.5k | Open | 43 | $0.18 | 139 | 0.74 | 18.77 | 14.42 | |||
Inkling (FP8) | 131k | Open | 42 | $0.43 | 152 | 0.58 | 16.99 | 13.13 | |||
 Hy3 (FP8) | 262k | Open | 42 | $0.04 | 64 | 0.64 | 39.71 | 31.26 | |||
DeepSeek V4 Flash (max) (FP4) | 1.05M | Open | 42 | $0.04 | 37 | 1.57 | 166.20 | 151.17 | |||
Inkling Small | 524k | Open | 41 | $0.15 | 124 | 0.82 | 20.97 | 16.13 | |||
GLM-5.1 (FP4) | 203k | Open | 41 | $0.19 | 27 | 0.89 | 161.63 | 142.01 | |||
GLM-5 (FP4) | 203k | Open | 41* | -- | 38 | 0.85 | 96.43 | 82.33 | |||
DeepSeek V4 Flash (high) (FP4) | 1.05M | Open | 39 | $0.03 | 36 | 1.29 | 49.21 | 34.15 | |||
Nemotron 3 Ultra | 262k | Open | 38 | $0.23 | 101 | 5.92 | 33.48 | 22.59 | |||
Nemotron 3 Ultra BF16 | 262k | Open | 38 | -- | 83 | 4.55 | 37.88 | 27.32 | |||
MiMo-V2.5 | 262k | Open | 38 | $0.08 | 37 | 1.02 | 67.99 | 53.58 | |||
Ling 3.0 Flash | 131k | Open | 38 | $0.04 | 289 | 0.64 | 9.28 | 6.91 | |||
 Qwen3.6 27B FP8 | 262k | Open | 38 | $0.19 | 50 | 0.98 | 124.06 | 113.11 | |||
GLM-5.1 (FP4) | 203k | Open | 36* | -- | 29 | 0.97 | 17.93 | -- | |||
Kimi K2.5 | 262k | Open | 36 | $0.07 | 46 | 0.74 | 75.45 | 63.94 | |||
Kimi K2.6 (FP4) | 262k | Open | 35* | -- | 16 | 1.71 | 33.47 | -- | |||
Muse Glimmer (high) | 131k | Open | 35 | $0.07 | 139 | 0.99 | 19.03 | 14.43 | |||
Qwen3.5 27B (FP8) | 262k | Open | 35* | -- | 65 | 0.96 | 39.37 | 30.73 | |||
MiniMax-M2.5 (FP8) | 197k | Open | 34* | -- | 29 | 1.52 | 87.00 | 68.39 | |||
GLM-4.7 (FP4) | 203k | Open | 34 | $0.10 | 39 | 0.96 | 65.51 | 51.64 | |||
Qwen3.5 397B A17B (FP8) | 262k | Open | 34 | $0.19 | 47 | 0.95 | 79.41 | 67.82 | |||
GLM-5 (FP8) | 203k | Open | 33* | -- | 31 | 1.07 | 17.18 | -- | |||
Qwen3.5 122B A10B (FP4) | 262k | Open | 33 | $0.12 | 124 | 0.74 | 20.97 | 16.18 | |||
DeepSeek V3.2 (FP4) | 164k | Open | 33* | -- | 22 | 0.76 | 114.11 | 90.68 | |||
Qwen3.5 397B A17B (FP8) | 262k | Open | 33* | -- | 44 | 0.91 | 12.19 | -- | |||
Qwen3.6 35B A3B (FP8) | 262k | Open | 32 | $0.09 | 56 | 0.79 | 106.01 | 96.29 | |||
Qwen3.6 27B FP8 | 262k | Open | 31 | $0.23 | 49 | 1.26 | 11.39 | -- | |||
Step 3.7 Flash | 256k | Open | 31 | -- | 176 | 0.69 | 14.86 | 11.34 | |||
Qwen3.5 27B FP8 | 262k | Open | 30* | -- | 71 | 0.98 | 8.04 | -- | |||
Qwen3.5 35B A3B (FP8) | 262k | Open | 30* | -- | 142 | 0.48 | 18.09 | 14.09 | |||
Gemma 4 31B | 262k | Open | 30 | $0.01 | 19 | 2.27 | 118.90 | 90.55 | |||
GLM-4.6 (FP4) | 203k | Open | 29 | -- | 35 | 1.04 | 71.80 | 56.61 | |||
MiMo-V2.5-Pro | 65.5k | Open | 28* | -- | 138 | 0.73 | 4.35 | -- | |||
Qwen3.5 122B A10B (FP4) | 262k | Open | 28 | $0.10 | 115 | 0.83 | 5.17 | -- | |||
GLM-4.7 (FP4) | 203k | Open | 27* | -- | 41 | 1.08 | 13.17 | -- | |||
Gemma 4 26B A4B | 262k | Open | 26 | $0.03 | 22 | 0.88 | 114.20 | 90.66 | |||
Nemotron 3 Super | 262k | Open | 26 | $0.10 | 69 | 11.13 | 47.23 | 28.88 | |||
DeepSeek V3.2 | 164k | Open | 25* | -- | 17 | 0.92 | 29.98 | -- | |||
Qwen3.6 35B A3B (FP8) | 262k | Open | 25* | -- | 69 | 0.74 | 8.00 | -- | |||
Qwen3.5 35B A3B FP8 | 262k | Open | 24 | $0.07 | 147 | 0.51 | 3.90 | -- | |||
gpt-oss-120b (high) | 131k | Open | 24 | $0.02 | 42 | 0.92 | 60.60 | 47.74 | |||
gpt-oss-120b (high) (Turbo) | 131k | Open | 24 | $0.07 | 159 | 0.72 | 16.41 | 12.55 | |||
Nemotron 3.5 Lightning (NVFP4) | 28.7k | Open | 24 | $0.06 | 666 | 0.49 | 4.24 | 3.01 | |||
GLM-4.7-Flash | 203k | Open | 23* | -- | 49 | 0.78 | 52.28 | 41.20 | |||
Gemma 4 31B (FP8) | 262k | Open | 22 | $0.04 | 16 | 1.48 | 32.08 | -- | |||
DeepSeek V3.1 Terminus (FP4) | 164k | Open | 22* | -- | 67 | 0.79 | 8.27 | -- | |||
DeepSeek V3.1 (FP4) | 164k | Open | 21* | -- | 12 | 1.20 | 42.82 | -- | |||
Gemma 4 26B A4B (FP8) | 262k | Open | 20* | -- | 24 | 0.72 | 21.38 | -- | |||
Qwen3.5 4B (FP8) | 262k | Open | 20* | -- | 31 | 0.78 | 82.14 | 65.09 | |||
DeepSeek R1 0528 | 164k | Open | 20* | -- | 28 | 0.80 | 88.94 | 70.51 | |||
Qwen3 235B A22B 2507 (FP8) | 262k | Open | 20 | -- | 61 | 1.00 | 41.73 | 32.59 | |||
Qwen3 235B 2507 (FP8) | 262k | Open | 18* | -- | 20 | 0.62 | 25.13 | -- | |||
Qwen3 Coder 480B (Turbo, FP4) | 262k | Open | 18* | -- | 60 | 0.85 | 9.14 | -- | |||
Qwen3.5 4B FP8 | 262k | Open | 16* | -- | 32 | 0.77 | 16.21 | -- | |||
DeepSeek V3 0324 (FP4) | 164k | Open | 15 | $0.03 | 31 | 2.02 | 18.23 | -- | |||
gpt-oss-20b (high) | 131k | Open | 15 | $0.01 | 95 | 0.46 | 26.70 | 21.00 | |||
 Mistral Small 3.1 | 128k | Open | 15 | $0.02 | 34 | 1.03 | 15.62 | -- | |||
Nemotron 3 Nano | 262k | Open | 15 | $0.02 | 133 | 3.80 | 22.58 | 15.03 | |||
Llama 4 Maverick (FP8) | 1.05M | Open | 14 | $0.03 | 87 | 0.62 | 6.35 | -- | |||
DeepSeek V3 (Dec) | 164k | Open | 14 | $0.02 | 25 | 0.83 | 20.79 | -- | |||
Qwen3 Next 80B A3B | 262k | Open | 14* | -- | 101 | 0.67 | 5.63 | -- | |||
Llama Nemotron Super 49B v1.5 | 131k | Open | 12* | -- | 81 | 4.34 | 35.11 | 24.62 | |||
Gemma 4 E4B | 262k | Open | 12* | -- | 72 | 0.92 | 35.62 | 27.76 | |||
Qwen3 32B (FP8) | 41k | Open | 11 | -- | 60 | 0.71 | 42.16 | 33.16 | |||
Mistral Small 3.2 (FP8) | 128k | Open | 11 | $0.11 | 29 | 0.88 | 17.86 | -- | |||
Qwen3 14B (FP8) | 32.8k | Open | 10 | -- | 40 | 0.94 | 63.72 | 50.22 | |||
Llama 4 Scout | 328k | Open | 10 | $0.01 | 53 | 0.72 | 10.22 | -- | |||
DeepSeek R1 Distill Llama 70B | 131k | Open | 10* | -- | 28 | 0.81 | 89.50 | 70.96 | |||
Qwen2.5 72B | 32.8k | Open | 9* | -- | 25 | 7.94 | 28.06 | -- | |||
Llama 3.3 70B (Turbo, FP8) | 131k | Open | 9* | -- | 13 | 1.97 | 41.64 | -- | |||
Qwen3 30B (FP8) | 41k | Open | 9* | -- | 82 | 0.51 | 31.01 | 24.40 | |||
NVIDIA Nemotron Nano 12B v2 VL (FP8) | 131k | Open | 9* | -- | 84 | 4.62 | 34.27 | 23.72 | |||
Gemma 4 E4B | 262k | Open | 9* | -- | 72 | 0.78 | 7.74 | -- | |||
NVIDIA Nemotron Nano 9B V2 | 131k | Open | 9* | -- | 131 | 5.32 | 24.41 | 15.27 | |||
Llama Nemotron Super 49B v1.5 | 131k | Open | 9* | -- | 102 | 7.33 | 12.22 | -- | |||
Qwen3 32B (FP8) | 41k | Open | 8* | -- | 60 | 0.69 | 9.04 | -- | |||
Llama 3.1 Nemotron 70B | 131k | Open | 7* | -- | 91 | 5.01 | 10.49 | -- | |||
Llama 3.1 8B (Turbo, FP8) | 131k | Open | 7* | -- | 17 | 1.09 | 30.85 | -- | |||
Llama 3.1 8B | 131k | Open | 7* | -- | 17 | 1.13 | 30.76 | -- | |||
Gemma 3 27B | 131k | Open | 7 | $0.08 | 26 | 1.18 | 20.71 | -- | |||
Nemotron 3 Nano | 262k | Open | 7* | -- | 130 | 0.40 | 4.25 | -- | |||
NVIDIA Nemotron Nano 9B V2 | 131k | Open | 7* | -- | 127 | 4.70 | 8.63 | -- | |||
Qwen3 14B (FP8) | 41k | Open | 7* | -- | 38 | 0.88 | 13.89 | -- | |||
Mistral Small 3 | 32.8k | Open | 7* | -- | 50 | 0.87 | 10.78 | -- | |||
Qwen3 30B (FP8) | 41k | Open | 7* | -- | 79 | 0.52 | 6.85 | -- | |||
Llama 3.1 70B | 131k | Open | 7* | -- | 28 | 1.92 | 19.76 | -- | |||
Llama 3.1 70B (Turbo, FP8) | 131k | Open | 7* | -- | 30 | 1.85 | 18.73 | -- | |||
Gemma 3 12B | 131k | Open | 6 | $0.09 | 37 | 0.92 | 14.38 | -- | |||
 Hermes 3 - Llama-3.1 70B | 131k | Open | 5* | -- | 32 | 1.99 | 17.71 | -- | |||
 Phi-4 | 16.4k | Open | 5* | -- | 74 | 0.92 | 7.70 | -- | |||
NVIDIA Nemotron Nano 12B v2 VL (FP8) | 131k | Open | 4* | -- | 78 | 6.55 | 12.99 | -- | |||
Llama 3.2 11B (Vision) | 131k | Open | 3* | -- | 23 | 1.84 | 23.47 | -- | |||
Gemma 3 4B | 131k | Open | 1* | -- | 40 | 0.90 | 13.28 | -- | |||
Llama 3 8B | 8.19k | Open | 1* | -- | -- | -- | -- | -- | |||
Key definitions
Frequently Asked Questions
Common questions about DeepInfra
DeepInfra offers 100 models that we track: GLM-5.2 (max) (FP4), DeepSeek V4 Flash 0731 (max), MiniMax-M3, DeepSeek V4 Pro (max) (FP4), Kimi K2.6 (FP4), DeepSeek V4 Pro (high) (FP4), Kimi K2.7 Code, MiMo-V2.5-Pro, Inkling (FP8), Hy3 (FP8), DeepSeek V4 Flash (max) (FP4), Inkling Small, GLM-5.1 (FP4), GLM-5 (FP4), DeepSeek V4 Flash (high) (FP4), Nemotron 3 Ultra, Nemotron 3 Ultra BF16, MiMo-V2.5, Ling 3.0 Flash, Qwen3.6 27B FP8, GLM-5.1 (FP4), Kimi K2.5, Kimi K2.6 (FP4), Muse Glimmer (high), Qwen3.5 27B (FP8), MiniMax-M2.5 (FP8), GLM-4.7 (FP4), Qwen3.5 397B A17B (FP8), GLM-5 (FP8), Qwen3.5 122B A10B (FP4), DeepSeek V3.2 (FP4), Qwen3.5 397B A17B (FP8), Qwen3.6 35B A3B (FP8), Qwen3.6 27B FP8, Step 3.7 Flash, Qwen3.5 27B FP8, Qwen3.5 35B A3B (FP8), Gemma 4 31B, GLM-4.6 (FP4), MiMo-V2.5-Pro, Qwen3.5 122B A10B (FP4), GLM-4.7 (FP4), Gemma 4 26B A4B, Nemotron 3 Super, DeepSeek V3.2, Qwen3.6 35B A3B (FP8), Qwen3.5 35B A3B FP8, gpt-oss-120b (high), gpt-oss-120b (high) (Turbo), Nemotron 3.5 Lightning (NVFP4), GLM-4.7-Flash, Gemma 4 31B (FP8), DeepSeek V3.1 Terminus (FP4), DeepSeek V3.1 (FP4), Gemma 4 26B A4B (FP8), Qwen3.5 4B (FP8), DeepSeek R1 0528, Qwen3 235B A22B 2507 (FP8), Qwen3 235B 2507 (FP8), Qwen3 Coder 480B (Turbo, FP4), Qwen3.5 4B FP8, DeepSeek V3 0324 (FP4), gpt-oss-20b (high), Mistral Small 3.1, Nemotron 3 Nano, Llama 4 Maverick (FP8), DeepSeek V3 (Dec), Qwen3 Next 80B A3B, Llama Nemotron Super 49B v1.5, Gemma 4 E4B, Qwen3 32B (FP8), Mistral Small 3.2 (FP8), Qwen3 14B (FP8), Llama 4 Scout, DeepSeek R1 Distill Llama 70B, Qwen2.5 72B, Llama 3.3 70B (Turbo, FP8), Qwen3 30B (FP8), NVIDIA Nemotron Nano 12B v2 VL (FP8), Gemma 4 E4B, NVIDIA Nemotron Nano 9B V2, Llama Nemotron Super 49B v1.5, Qwen3 32B (FP8), Llama 3.1 Nemotron 70B, Llama 3.1 8B (Turbo, FP8), Llama 3.1 8B, Gemma 3 27B, Nemotron 3 Nano, NVIDIA Nemotron Nano 9B V2, Qwen3 14B (FP8), Mistral Small 3, Qwen3 30B (FP8), Llama 3.1 70B, Llama 3.1 70B (Turbo, FP8), Gemma 3 12B, Hermes 3 - Llama-3.1 70B, Phi-4, NVIDIA Nemotron Nano 12B v2 VL (FP8), Llama 3.2 11B (Vision), and Gemma 3 4B.
The most intelligent model available on DeepInfra is GLM-5.2 (max) (FP4) with an Intelligence Index score of 53.
The fastest model on DeepInfra by output speed is Nemotron 3.5 Lightning (NVFP4) at 665.5 tokens per second.
The model with the lowest time to first answer token on DeepInfra is Nemotron 3 Nano at 0.40s. Lower latency means faster initial response time.
The most affordable model on DeepInfra by blended price is Llama 3.1 8B (Turbo, FP8) at $0.02 per 1M tokens (7:2:1 cache hit/input/output ratio).
Prices on DeepInfra vary up to 55x across models, from $0.02 per 1M tokens for Llama 3.1 8B (Turbo, FP8) to $1.20 per 1M tokens for Llama 3.1 Nemotron 70B.
Yes, DeepInfra offers an OpenAI-compatible API, making it easy to switch from OpenAI or use existing OpenAI SDK integrations.
96 of 100 models on DeepInfra support JSON mode for structured output.
96 of 100 models on DeepInfra support function calling (tool use).
Yes, DeepInfra offers 52 reasoning models: GLM-5.2 (max) (FP4), DeepSeek V4 Flash 0731 (max), MiniMax-M3, DeepSeek V4 Pro (max) (FP4), Kimi K2.6 (FP4), DeepSeek V4 Pro (high) (FP4), Kimi K2.7 Code, MiMo-V2.5-Pro, Inkling (FP8), Hy3 (FP8), DeepSeek V4 Flash (max) (FP4), Inkling Small, GLM-5.1 (FP4), GLM-5 (FP4), DeepSeek V4 Flash (high) (FP4), Nemotron 3 Ultra, Nemotron 3 Ultra BF16, MiMo-V2.5, Ling 3.0 Flash, Qwen3.6 27B FP8, Kimi K2.5, Muse Glimmer (high), Qwen3.5 27B (FP8), MiniMax-M2.5 (FP8), GLM-4.7 (FP4), Qwen3.5 397B A17B (FP8), Qwen3.5 122B A10B (FP4), DeepSeek V3.2 (FP4), Qwen3.6 35B A3B (FP8), Step 3.7 Flash, Qwen3.5 35B A3B (FP8), Gemma 4 31B, GLM-4.6 (FP4), Gemma 4 26B A4B, Nemotron 3 Super, gpt-oss-120b (high), gpt-oss-120b (high) (Turbo), Nemotron 3.5 Lightning (NVFP4), GLM-4.7-Flash, Qwen3.5 4B (FP8), DeepSeek R1 0528, Qwen3 235B A22B 2507 (FP8), gpt-oss-20b (high), Nemotron 3 Nano, Llama Nemotron Super 49B v1.5, Gemma 4 E4B, Qwen3 32B (FP8), Qwen3 14B (FP8), DeepSeek R1 Distill Llama 70B, Qwen3 30B (FP8), NVIDIA Nemotron Nano 12B v2 VL (FP8), and NVIDIA Nemotron Nano 9B V2. Reasoning models use extended thinking to work through complex problems before providing an answer.
Yes, all 100 models on DeepInfra are open weight models.
Yes, provider performance can vary over time due to infrastructure changes, load balancing, and updates. We continuously benchmark all providers and display historical performance trends in the "Over Time" charts.
When choosing a model on DeepInfra, consider: intelligence (for quality-sensitive tasks), output speed (for throughput-intensive tasks), latency (for interactive applications requiring quick first responses), pricing (for cost-sensitive workloads), and features like context window size, JSON mode, or function calling support.