Speed & Latency
Output speed (tokens per second) and latency (time to first token) benchmarks across models and providers.
Filters:
Provider:
Creators:
Speed vs Intelligence
EUGlobal
Drag to zoom ยท Click to pin69 models
| # | Model | Creator | Output Speed (t/s)โ | Latency TTFT (s) | Total Response (s) | Intelligence | Price $/1M |
|---|---|---|---|---|---|---|---|
| 1 | GPT-OSS 120BOpenEU | OpenAI | 362.9 t/s | 0.52s | - | 33.3 | $0.26 |
| 2 | Gemini 2.5 Flash-LiteEU | 311.4 t/s | 22.36s | - | 17.6 | $0.17 | |
| 3 | Gemini 3.1 Flash-LiteGlobal | 296.3 t/s | 5.38s | - | 33.5 | $0.56 | |
| 4 | o3-miniREU | OpenAI | 212.4 t/s | 20.09s | - | 25.2 | $1.93 |
| 5 | Gemini 2.5 FlashREU | 212.3 t/s | 10.21s | - | 27 | $0.85 | |
| 6 | Gemini 3.5 FlashRGlobal | 212.1 t/s | 12.39s | - | 55.3 | $3.38 | |
| 7 | MiniMax M2.1REU | MiniMax | 210.5 t/s | 2.36s | - | 39.4 | $0.53 |
| 8 | MiniMax M2.5RGlobal | MiniMax | 202.4 t/s | 2.38s | - | 41.9 | $0.53 |
| 9 | Gemini 3 FlashRGlobal | 184.5 t/s | 5.29s | - | 46.4 | $1.13 | |
| 10 | Grok 4.3RGlobal | xAI | 176.4 t/s | 21.45s | - | 53.2 | $1.56 |
| 11 | Mistral Small 4ROpenGlobal | Mistral | 176.2 t/s | 0.50s | - | 27.8 | $0.26 |
| 12 | Grok 4.20RGlobal | xAI | 170.6 t/s | 15.73s | - | 49.3 | $3.00 |
| 13 | GPT-5 NanoREU | OpenAI | 168.8 t/s | 88.92s | - | 26.8 | $0.14 |
| 14 | GPT-5.4 NanoRGlobal | OpenAI | 161.3 t/s | 3.48s | - | 44 | $0.46 |
| 15 | GPT-5.4 MiniRGlobal | OpenAI | 158.7 t/s | 4.17s | - | 48.9 | $1.69 |
| 16 | o4-miniREU | OpenAI | 153.9 t/s | 17.65s | - | 33.1 | $1.93 |
| 17 | Nemotron 3 SuperROpenGlobal | NVIDIA | 150.2 t/s | 1.08s | - | 36 | $0.41 |
| 18 | GPT-4oEU | OpenAI | 142.1 t/s | 0.50s | - | 17.3 | $4.38 |
| 19 | Nova 2 LiteREU | Amazon | 139.7 t/s | 17.29s | - | 34.5 | $0.85 |
| 20 | Claude Haiku 4.5REU | Anthropic | 135.4 t/s | 11.09s | - | 37.1 | $2.19 |
| 21 | GPT-4.1EU | OpenAI | 128.7 t/s | 0.55s | - | 26.3 | $3.50 |
| 22 | Gemini 2.5 ProREU | 123.6 t/s | 19.28s | - | 34.6 | $3.44 | |
| 23 | Gemini 3.1 ProRGlobal | 123.3 t/s | 20.05s | - | 57.2 | $4.50 | |
| 24 | Mistral Medium 3.5Global | Mistral | 120.6 t/s | 0.75s | - | 39.2 | $3.00 |
| 25 | GPT-5.1REU | OpenAI | 118.6 t/s | 29.12s | - | 47.7 | $3.44 |
| 26 | DeepSeek V4 FlashREU | DeepSeek | 118.6 t/s | 0.88s | - | 46.5 | $0.17 |
| 27 | o1REU | OpenAI | 116.3 t/s | 15.39s | - | 30.7 | $26.25 |
| 28 | GPT-4.1 NanoEU | OpenAI | 114.4 t/s | 0.42s | - | 13 | $0.17 |
| 29 | Llama 4 MaverickOpenGlobal | Meta | 113.7 t/s | 0.61s | - | 18.4 | $0.47 |
| 30 | o3REU | OpenAI | 106.5 t/s | 6.69s | - | 38.4 | $3.50 |
| 31 | GPT-5REU | OpenAI | 105.4 t/s | 84.61s | - | 44.6 | $3.44 |
| 32 | Llama 3.3 70BOpenGlobal | Meta | 92.3 t/s | 0.64s | - | 14.5 | $0.61 |
| 33 | GPT-5 MiniREU | OpenAI | 89.9 t/s | 78.72s | - | 41.2 | $0.69 |
| 34 | GPT-4.1 MiniEU | OpenAI | 89.2 t/s | 0.50s | - | 22.9 | $0.70 |
| 35 | Hermes 4 70BROpenEU | NousResearch | 87.4 t/s | 0.64s | - | 16 | $0.20 |
| 36 | GLM-5RGlobal | Z AI | 81.4 t/s | 0.82s | - | 49.8 | $1.55 |
| 37 | GPT-5.4RGlobal | OpenAI | 80.3 t/s | 195.49s | - | 56.8 | $5.63 |
| 38 | GPT-5.2RGlobal | OpenAI | 71 t/s | 89.53s | - | 51.3 | $4.81 |
| 39 | GPT-5.3 CodexRGlobal | OpenAI | 70.6 t/s | 64.99s | - | 53.6 | $4.81 |
| 40 | GPT-4o MiniEU | OpenAI | 69.8 t/s | 0.57s | - | 12.6 | $0.26 |
| 41 | MiniMax M2.7RGlobal | MiniMax | 66.6 t/s | 2.94s | - | 49.6 | $0.53 |
| 42 | Claude Sonnet 4.6REU | Anthropic | 64.3 t/s | 67.87s | - | 51.7 | $6.56 |
| 43 | Grok 3 MiniRGlobal | xAI | 63.9 t/s | 0.61s | - | 32.1 | $0.35 |
| 44 | Qwen3 235BROpenEU | Alibaba | 61.7 t/s | 1.24s | - | 19.8 | $2.63 |
| 45 | GPT-5.5RGlobal | OpenAI | 61.7 t/s | 70.44s | - | 60.2 | $11.25 |
| 46 | Mistral Large 3OpenGlobal | Mistral | 60.6 t/s | 0.57s | - | 22.8 | $0.75 |
| 47 | Claude Opus 4.8REU | Anthropic | 57.5 t/s | 25.68s | - | 61.4 | $10.94 |
| 48 | Claude Opus 4.5REU | Anthropic | 54.3 t/s | 12.51s | - | 49.7 | $10.94 |
| 49 | Qwen3.6 PlusRGlobal | Alibaba | 52.7 t/s | 1.84s | - | 50 | $1.13 |
| 50 | Claude Opus 4.7REU | Anthropic | 52.1 t/s | 15.94s | - | 57.3 | $10.94 |
| 51 | Qwen3.5 397BROpenGlobal | Alibaba | 51.5 t/s | 1.81s | - | 45 | $1.35 |
| 52 | GLM-5.1RGlobal | Z AI | 50.8 t/s | 1.01s | - | 51.4 | $2.15 |
| 53 | Claude Opus 4.6REU | Anthropic | 49.1 t/s | 13.49s | - | 52.9 | $10.94 |
| 54 | Claude Sonnet 4.5REU | Anthropic | 48.9 t/s | 9.16s | - | 43 | $6.56 |
| 55 | DeepSeek V4 ProREU | DeepSeek | 48.2 t/s | 1.18s | - | 51.5 | $0.54 |
| 56 | Claude Sonnet 4REU | Anthropic | 46.7 t/s | 9.60s | - | 38.7 | $6.56 |
| 57 | MiMo V2 ProROpenGlobal | Xiaomi | 42.5 t/s | 2.65s | - | 49.2 | $1.50 |
| 58 | Hermes 4 405BROpenEU | NousResearch | 41.1 t/s | 0.77s | - | 18.6 | $1.50 |
| 59 | Kimi K2.5RGlobal | Kimi | 33.3 t/s | 1.32s | - | 46.8 | $1.19 |
| 60 | Gemini 3 ProRGlobal | - | - | - | 48.4 | $4.50 | |
| 61 | Muse SparkRGlobal | Meta | - | - | - | 52.2 | $0.00 |
| 62 | Grok 4RGlobal | xAI | - | - | - | 41.5 | $11.00 |
| 63 | DeepSeek V3.2ROpenGlobal | DeepSeek | - | - | - | 41.7 | $0.34 |
| 64 | DeepSeek V3.2 (Non-reasoning)OpenGlobal | DeepSeek | - | - | - | 32.1 | $0.78 |
| 65 | Magistral MediumRGlobal | Mistral | - | - | - | 18.8 | $1.00 |
| 66 | Magistral SmallRGlobal | Mistral | - | - | - | 16.8 | $0.50 |
| 67 | GPT-5.4 ProRGlobal | OpenAI | - | - | - | - | $67.50 |
| 68 | GPT-5.5 ProRGlobal | OpenAI | - | - | - | - | $0.00 |
| 69 | Grok 3Global | xAI | - | - | - | 25.2 | $8.00 |