☁️ Cloud Benchmarks ☁️
I run cron jobs to periodically test the token generation speed of different cloud LLM providers. The chart helps visualize the distributions of different speeds, as they can vary somewhat depending on the loads. For readability not all models are shown, but you can see the full results in the table below.
Every provider and model now has a dedicated landing page with narrative insights, SEO-friendly metadata, and structured data for search engines. Click any provider or model in the table to explore performance in depth.
I am working daily to add more providers and models, looking anywhere that does not require purchasing dedicated endpoints for hosting (why some models may appear to be missing). If you have any more suggestions let me know on GitHub!! 😊
Pick A Path In 10 Seconds
Quick recommendations from the latest 7-day benchmark slice. Use one path, jump into full results, then drill into provider/model pages.
Loading 7-day recommendations…
Fastest Models Right Now (updated <24h)
| # | Model | Provider | Speed |
|---|---|---|---|
| 1 | llama-3.1-8b | groq | 318 tok/s |
| 2 | qwen-3-32b | groq | 229 tok/s |
| 3 | llama-4-scout | groq | 208 tok/s |
| 4 | llama-3.3-70b | groq | 202 tok/s |
| 5 | llama-4-maverick | groq | 196 tok/s |
📊 Speed Distribution 📊
📚 Full Results 📚
Showing 78 of 78 modelsFlagged statuses: likely_deprecated, deprecated, failing, stale, never_succeeded, disabled
| Status | |||||||
|---|---|---|---|---|---|---|---|
| groq | llama-3.1-8b | Active | 36m ago | 318.00 | 87 | 471 | 90.00 |
| cerebras | qwen-3-32b | Active | 21d ago | 260.00 | 32 | 444 | 240.00 |
| groq | qwen-3-32b | Active | 36m ago | 229.00 | 2 | 391 | 300.00 |
| groq | llama-4-scout | Active | 36m ago | 208.00 | 38 | 335 | 220.00 |
| groq | llama-3.3-70b | Active | 36m ago | 202.00 | 68 | 340 | 130.00 |
| groq | llama-4-maverick | Active | 36m ago | 196.00 | 1 | 307 | 610.00 |
| cerebras | gpt-oss-120b | Active | 37m ago | 192.00 | 13 | 380 | 840.00 |
| cerebras | llama-3.1-8b | Active | 37m ago | 192.00 | 1 | 348 | 1000.00 |
| cerebras | llama-3.3-70b | Active | 21d ago | 185.00 | 17 | 316 | 400.00 |
| together | llama-3.1-8b | Active | 34m ago | 142.00 | 3 | 232 | 380.00 |
| groq | kimi-k2 | Active | 36m ago | 141.00 | 12 | 215 | 310.00 |
| bedrock | nova-micro | Active | 19m ago | 122.00 | 65 | 152 | 270.00 |
| openai | o3 Mini | Never Succeeded(Medium) | 35m ago | 111.00 | 21 | 160 | 0.00 |
| bedrock | llama-4-maverick | Active | 19m ago | 108.00 | 3 | 139 | 260.00 |
| bedrock | nova-lite | Active | 19m ago | 102.00 | 39 | 132 | 300.00 |
| bedrock | llama-4-scout | Active | 19m ago | 101.00 | 4 | 129 | 280.00 |
| bedrock | llama-3.3-70b | Active | 19m ago | 98.40 | 3 | 136 | 290.00 |
| together | qwen-2.5-7b | Active | 34m ago | 94.20 | 11 | 145 | 240.00 |
| bedrock | nova-pro | Active | 19m ago | 90.20 | 31 | 121 | 350.00 |
| openai | GPT-5.1-codex-max | Active | 35m ago | 80.50 | 11 | 118 | 1420.00 |
| openai | gpt-3.5-turbo | Active | 34m ago | 78.20 | 13 | 126 | 470.00 |
| gemini-2.5-flash-lite | Active | 34m ago | 76.00 | 10 | 132 | 520.00 | |
| openai | gpt-4.1-nano | Active | 35m ago | 72.30 | 9 | 149 | 450.00 |
| together | llama-3.1-70b | Active | 11d ago | 72.10 | 7 | 129 | 390.00 |
| together | mistral-7b | Active | 11d ago | 71.00 | 2 | 91 | 480.00 |
| deepinfra | mistral-7b | Stale(Medium) | 36m ago | 70.10 | 5 | 148 | 620.00 |
| openai | gpt-4o | Active | 34m ago | 69.70 | 13 | 173 | 1310.00 |
| fireworks | mixtral-8x22b | Active | 7d ago | 67.00 | 29 | 112 | 440.00 |
| gemini-2.5-flash | Never Succeeded(Medium) | 34m ago | 66.90 | 5 | 101 | 1030.00 | |
| deepinfra | devstral-small | Never Succeeded(Medium) | 37m ago | 64.60 | 9 | 140 | 580.00 |
| together | mixtral-8x7b | Active | 34m ago | 61.20 | 14 | 114 | 150.00 |
| deepinfra | mixtral-8x22b | Stale(Medium) | 19d ago | 59.70 | 14 | 78 | 420.00 |
| together | llama-3.2-3b | Active | 2d ago | 57.30 | 5 | 121 | 1510.00 |
| together | llama-3.3-70b | Active | 34m ago | 54.00 | 1 | 146 | 1350.00 |
| fireworks | llama-3.3-70b | Active | 36m ago | 51.80 | 2 | 95 | 1380.00 |
| anthropic | claude-haiku-4.5 | Active | 37m ago | 51.70 | 19 | 74 | 550.00 |
| openai | gpt-4.1-mini | Active | 35m ago | 51.40 | 15 | 85 | 450.00 |
| openai | o4 Mini | Never Succeeded(Medium) | 35m ago | 49.80 | 14 | 76 | 0.00 |
| deepinfra | llama-3.1-8b | Stale(Medium) | 36m ago | 47.40 | 4 | 96 | 570.00 |
| bedrock | llama-3.2-90b | Active | 19m ago | 47.10 | 2 | 51 | 350.00 |
| together | deepseek-r1 | Active | 34m ago | 46.80 | 1 | 113 | 930.00 |
| deepinfra | llama-3.2-1b | Stale(Medium) | 36m ago | 46.70 | 1 | 100 | 610.00 |
| deepinfra | llama-3.2-3b | Stale(Medium) | 36m ago | 45.80 | 2 | 99 | 520.00 |
| deepinfra | llama-3-8b | Stale(Medium) | 36m ago | 45.80 | 9 | 71 | 310.00 |
| openai | gpt-4o-mini | Active | 34m ago | 42.00 | 20 | 95 | 360.00 |
| bedrock | mistral-large | Active | 19m ago | 41.90 | 3 | 47 | 390.00 |
| bedrock | claude-haiku-4.5 | Active | 19m ago | 41.50 | 4 | 62 | 1090.00 |
| gemini-2.5-pro | Never Succeeded(Medium) | 34m ago | 41.20 | 2 | 72 | 1680.00 | |
| openai | gpt-4.1 | Active | 35m ago | 38.60 | 6 | 83 | 500.00 |
| deepinfra | llama-3.2-90b | Stale(Medium) | 36m ago | 35.30 | 3 | 82 | 870.00 |
| deepinfra | llama-2-70b | Stale(Medium) | 36m ago | 34.80 | 3 | 57 | 580.00 |
| deepinfra | llama-3-70b | Stale(Medium) | 36m ago | 34.30 | 2 | 55 | 680.00 |
| bedrock | claude-3-7-sonnet | Active | 19m ago | 33.20 | 2 | 44 | 730.00 |
| openai | gpt-4-turbo | Active | 34m ago | 32.60 | 2 | 52 | 530.00 |
| bedrock | claude-3-5-sonnet | Active | 19m ago | 32.40 | 2 | 44 | 660.00 |
| deepinfra | qwen-2.5-72b | Stale(Medium) | 36m ago | 32.40 | 1 | 50 | 760.00 |
| deepinfra | Qwen 2.5 Coder 32B | Never Succeeded(Medium) | 37m ago | 31.40 | 1 | 82 | 2710.00 |
| bedrock | claude-3-5-haiku | Active | 19m ago | 30.90 | 4 | 38 | 680.00 |
| openai | GPT-5.1 | Active | 35m ago | 30.10 | 3 | 57 | 950.00 |
| openai | gpt-4 | Active | 34m ago | 27.90 | 3 | 47 | 670.00 |
| openai | GPT-5.2 | Active | 35m ago | 26.90 | 9 | 40 | 910.00 |
| openai | GPT-5.1-codex | Active | 35m ago | 25.80 | 1 | 48 | 1290.00 |
| openai | GPT-5.1-codex-mini | Active | 35m ago | 25.30 | 1 | 55 | 1170.00 |
| deepinfra | llama-3.1-405b | Stale(Medium) | 36m ago | 25.20 | 1 | 39 | 970.00 |
| bedrock | claude-sonnet-4.5 | Active | 19m ago | 22.30 | 2 | 29 | 1620.00 |
| deepinfra | llama-3.1-70b | Stale(Medium) | 36m ago | 21.70 | 1 | 46 | 1120.00 |
| anthropic | claude-4-sonnet | Active | 37m ago | 20.40 | 1 | 31 | 1930.00 |
| anthropic | claude-opus-4.5 | Active | 37m ago | 20.10 | 2 | 33 | 1840.00 |
| bedrock | claude-3-opus | Active | 18d ago | 19.40 | 6 | 22 | 870.00 |
| deepinfra | llama-3.3-70b | Never Succeeded(Medium) | 37m ago | 19.10 | 1 | 46 | 2510.00 |
| bedrock | claude-opus-4.5 | Active | 19m ago | 18.40 | 1 | 24 | 2050.00 |
| anthropic | Claude Opus 4.1 | Active | 37m ago | 17.90 | 8 | 27 | 1490.00 |
| anthropic | claude-4-opus | Active | 37m ago | 17.60 | 5 | 24 | 1300.00 |
| openai | gpt-5.2-codex | Active | 6h ago | 12.90 | 1 | 27 | 1830.00 |
| deepinfra | llama-3.2-11b | Stale(Medium) | 36m ago | 12.80 | 1 | 62 | 2660.00 |
| deepinfra | qwen-3-235b | Never Succeeded(Medium) | 37m ago | 10.10 | 1 | 40 | 5950.00 |
| openai | o1-pro | Likely Deprecated(Medium) | 35m ago | 9.47 | 1 | 18 | 640.00 |
| openai | GPT-5.2-pro | Active | 36m ago | 8.53 | 4 | 14 | 4990.00 |
Lifecycle snapshot
Loading status summary…