LLM tokens per second — live multi-provider benchmark
Real inference speed, measured continuously. Every row is a live model from Ollama, OpenCode Zen, or OpenCode Go — sorted by tokens per second, benchmarked every ~10 minutes.
● live
— last benchmark 6s ago
| Trend 24h | ||||||||
|---|---|---|---|---|---|---|---|---|
| North Mini Code (Free) | OpenCode Zen | 494.4 | 329.6 | 3.1s | 63% | 20.6 | | 6m ago |
| deepseek-v4-flash | Ollama | 251.7 | 182.7 | 488ms | 100% | 40.3 | | 3m ago |
| MiMo V2.5 (Free) | OpenCode Zen | 216.8 | 313.7 | 3.9s | 75% | — | | 8m ago |
| nemotron-3-nano:30b (non-reasoning) | Ollama | 203.9 | 293.1 | 461ms | 100% | 7.4 | | 10m ago |
| deepseek-v4-pro | Ollama | 164.7 | 156.3 | 816ms | 100% | 44.3 | | 3m ago |
| kimi-k2.5 | Ollama | 157.1 | 128.0 | 908ms | 100% | 38.1 | | 51s ago |
| ministral-3:8b (non-reasoning) | Ollama | 149.0 | 132.1 | 448ms | 100% | 8.9 | | 11m ago |
| kimi-k2.7-code | Ollama | 148.8 | 138.0 | 890ms | 100% | 41.9 | | 38s ago |
| glm-5.2 | Ollama | 141.1 | 107.6 | 930ms | 100% | 50.7 | | 1m ago |
| gpt-oss:120b | Ollama | 128.1 | 145.0 | 458ms | 100% | 23.8 | | 1m ago |
| rnj-1:8b | Ollama | 126.7 | 126.3 | 397ms | 100% | — | | 8m ago |
| gemma4:31b | Ollama | 126.6 | 119.3 | 358ms | 100% | 29.4 | | 2m ago |
| glm-5 | Ollama | 124.4 | 98.4 | 1.1s | 100% | 39.5 | | 1m ago |
| qwen3.5:397b | Ollama | 120.3 | 96.5 | 567ms | 100% | 33.7 | | 8m ago |
| gemini-3-flash-preview | Ollama | 106.1 | 111.8 | 1.8s | 100% | 37.8 | | 2m ago |
| minimax-m2.1 | Ollama | 104.7 | 126.4 | — | 100% | 31.4 | | 32s ago |
| gpt-oss:20b | Ollama | 103.9 | 108.8 | 779ms | 100% | 14.9 | | 57s ago |
| kimi-k2.6 | Ollama | 102.7 | 94.3 | 653ms | 100% | 42.8 | | 45s ago |
| minimax-m3 | Ollama | 101.6 | 61.4 | 907ms | 100% | 44.4 | | 11m ago |
| qwen3-coder-next (non-reasoning) | Ollama | 99.2 | 92.4 | 344ms | 100% | 21.2 | | 9m ago |
| ministral-3:14b (non-reasoning) | Ollama | 97.2 | 102.5 | 457ms | 100% | 10 | | 11m ago |
| glm-5.1 | Ollama | 96.5 | 156.7 | 3.2s | 100% | 40.2 | | 1m ago |
| nemotron-3-super | Ollama | 94.4 | 56.1 | 463ms | 100% | 25.4 | | 10m ago |
| ministral-3:3b (non-reasoning) | Ollama | 91.1 | 187.0 | 487ms | 100% | 5.6 | | 11m ago |
| minimax-m2.5 | Ollama | 89.5 | 77.7 | 304ms | 100% | 33.7 | | 26s ago |
| glm-4.7 | Ollama | 88.3 | 87.8 | 1.6s | 100% | 33.8 | | 1m ago |
| devstral-2:123b (non-reasoning) | Ollama | 76.6 | 60.1 | 567ms | 100% | 15.5 | | 3m ago |
| devstral-small-2:24b (non-reasoning) | Ollama | 53.5 | 58.7 | 550ms | 100% | 13.1 | | 2m ago |
| gemma3:4b (non-reasoning) | Ollama | 51.1 | 51.1 | 539ms | 97% | 1.1 | | 2m ago |
| mistral-large-3:675b (non-reasoning) | Ollama | 47.7 | 58.5 | 629ms | 100% | 16.2 | | 11m ago |
| gemma3:12b (non-reasoning) | Ollama | 45.4 | 37.8 | 645ms | 100% | 3.4 | | 2m ago |
| minimax-m2.7 | Ollama | 37.2 | 39.5 | 1.2s | 100% | 38.1 | | 11m ago |
| qwen3-coder:480b (non-reasoning) | Ollama | 32.0 | 89.8 | 556ms | 100% | 18 | | 9m ago |
| gemma3:27b (non-reasoning) | Ollama | 18.5 | 16.0 | 613ms | 100% | 4.8 | | 2m ago |
| nemotron-3-ultra | Ollama | 12.0 | 17.8 | 19.2s | 99% | 37.8 | | 10m ago |
| deepseek-v3.1:671b (non-reasoning) | Ollama | 11.7 | 10.2 | 3.3s | 100% | 21 | | 5m ago |
| deepseek-v3.2 | Ollama | 3.9 | 25.4 | 1.0s | 99% | 33.4 | | 4m ago |
No models match your filter.
Intelligence Index scores from Artificial Analysis.