Ollama Cloud tokens per second — live benchmark
Real inference speed, measured continuously. Every row is a live Ollama Cloud model — sorted by tokens per second, benchmarked every ~10 minutes.
Weighing Ollama Cloud against a router? Ollama Cloud vs OpenRouter — what we measure, and what we don't.
| Trend | |||||||
|---|---|---|---|---|---|---|---|
| GPT-OSS 120B Free | 279.8 | 355.3 | 339ms | 100% | 11.6 | | 33m ago |
| Gemma4 31B Pro | 271.1 | 298.4 | 348ms | 100% | 19 | | 2m ago |
| Gemma4 31B Free | 222.1 | 148.4 | 325ms | 100% | 19 | | 39m ago |
| GPT-OSS 120B Pro | 217.4 | 223.7 | 375ms | 100% | 11.6 | | 6m ago |
| Nemotron 3 Nano 30B (non-reasoning) Free | 191.3 | 285.2 | 312ms | 100% | 6.8 | | 27m ago |
| Nemotron 3 Nano 30B (non-reasoning) Pro | 184.6 | 129.1 | 801ms | 100% | 6.8 | | 5m ago |
| DeepSeek V4 Flash 0731 Pro | 184.4 | 186.0 | 529ms | 100% | 34.3 | | 3m ago |
| DeepSeek V4.1 Flash Pro | 175.9 | 275.6 | 366ms | 99% | 39.5 | | 5m ago |
| DeepSeek V4 Pro 0813 Pro | 140.2 | 115.3 | 652ms | 100% | 36 | | 5m ago |
| GLM 5.3 Pro | 134.1 | 92.8 | 730ms | 100% | 44.8 | | 8m ago |
| GLM 5.3 Flash Pro | 117.4 | 93.0 | 710ms | 100% | 41.8 | | 4m ago |
| Kimi K2.7 Code Pro | 116.2 | 104.8 | 1.1s | 100% | 25.8 | | 3m ago |
| GPT-OSS 20B Free | 108.7 | 102.4 | 458ms | 100% | 9 | | 32m ago |
| GPT-OSS 20B Pro | 107.1 | 87.5 | 12.6s | 100% | 9 | | 6m ago |
| GLM 5.2 Pro | 106.2 | 166.0 | 651ms | 100% | 33.7 | | 8m ago |
| Qwen3.5 397B Pro | 91.6 | 95.0 | 1.0s | 99% | 18.4 | | 3m ago |
| Nemotron 3 Super Free | 91.0 | 63.1 | 541ms | 100% | 12.8 | | 15m ago |
| GLM 5.1 Pro | 90.0 | 102.1 | 1.0s | 100% | 26.1 | | 2m ago |
| Nemotron 3 Super Pro | 86.5 | 80.4 | 576ms | 99% | 12.8 | | 3m ago |
| MiniMax M3 Pro | 83.3 | 93.8 | 689ms | 99% | 29.2 | | 4m ago |
| Kimi K3 Pro | 81.0 | 77.2 | 853ms | 100% n=6 | 43.6 | | 2h ago |
| Mistral Large 3 675B (non-reasoning) Pro | 72.5 | 77.2 | 719ms | 100% | 9.3 | | 5m ago |
| MiniMax M2.7 Pro | 64.1 | 70.9 | 1.1s | 100% | 22.8 | | 5m ago |
| Kimi K2.6 Pro | 45.2 | 41.8 | 1.7s | 100% | 27 | | 6m ago |
| Nemotron 3 Ultra Free | 44.1 | 4.7 | 9.6s | 100% | 22.9 | | 59m ago |
| Nemotron 3 Ultra Pro | 40.5 | 64.7 | 568ms | 99% | 22.9 | | 11m ago |
| MiniMax M3 Free | — | 82.2 | 1.0s | 0% n=1 | 29.2 | | 21d ago |
No models match your filter.
Intelligence Index scores from Artificial Analysis.
Ollama Free is sampled about hourly to avoid burning through the weekly free-tier balance.
1 model unavailable or stale
- MiniMax M3 Free
Frequently asked questions
What is the fastest Ollama Cloud model right now?
As of the last build, GPT-OSS 120B is the fastest Ollama Cloud model on this leaderboard at 355.3 tokens/sec (most recent measured run). Rankings change as models are re-benchmarked roughly every 10 minutes (every 60 minutes on the Ollama Free tier) — see the live leaderboard above for the current order.
How is tokens per second measured?
TPS is generation throughput: output tokens divided by server-reported generation time, excluding time-to-first-token. For Ollama Cloud we use the server’s own reported timing (eval_count and total_duration) rather than a client-side stopwatch, so results are immune to network jitter and token buffering. Full formula and error handling are on the methodology page.
How often is data refreshed?
The worker benchmarks continuously using a round-robin priority queue. Ollama Cloud Pro is sampled about every 10 minutes; Ollama Free is sampled about every 60 minutes to preserve the weekly free-tier quota. Models Ollama bills as extra usage rather than plan usage — currently Kimi K3 — are sampled about every 4 hours, since each run is charged per token against a prepaid balance. The leaderboard also polls the API roughly every 60 seconds client-side to keep cells current without a full page reload.