Head-to-head The same models run on Ollama and OpenCode — who streams faster? See the race on TokenDyno →

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.

● live — last benchmark 9s ago
Trend
GPT-OSS 120B Free 324.9 290.2 320ms 100% 24.1 12m ago
GPT-OSS 120B Pro 223.8 168.0 1.1s 100% 24.1 7m ago
Nemotron 3 Nano 30B (non-reasoning) Pro 184.8 371.2 338ms 100% 7.2 14s ago
Nemotron 3 Nano 30B (non-reasoning) Free 181.1 114.6 581ms 100% 7.2 56m ago
Gemma4 31B Free 175.5 213.4 393ms 100% 29.7 13m ago
Gemma4 31B Pro 167.7 7.8 1.4s 98% 29.7 8m ago
DeepSeek V4 Flash 0731 Pro 166.5 74.1 688ms 100% 51.8 8m ago
DeepSeek V4 Pro 0813 Pro 156.5 155.4 434ms 100% 53.2 26s ago
GLM 5.3 Pro 128.1 86.2 712ms 100% 59.5 7m ago
GLM 5.3 Flash Pro 122.2 160.5 411ms 100% 57.5 9m ago
Kimi K2.7 Code Pro 119.9 142.1 679ms 100% 43 8m ago
MiniMax M3 Pro 110.7 192.4 438ms 100% 45.4 10m ago
GPT-OSS 20B Free 109.5 142.7 571ms 100% 15.2 11m ago
GPT-OSS 20B Pro 108.6 115.2 556ms 100% 15.2 7m ago
Kimi K3 Pro 107.5 115.3 687ms 100% n=6 59.7 3h ago
Nemotron 3 Super Free 99.0 20.3 595ms 100% 25.7 56m ago
Qwen3.5 397B Pro 92.3 79.3 1.0s 100% 34.3 9m ago
Nemotron 3 Super Pro 81.8 93.4 481ms 99% 25.7 9m ago
GLM 5.2 Pro 75.5 71.6 718ms 100% 52.6 7m ago
Mistral Large 3 675B (non-reasoning) Pro 69.7 85.3 600ms 99% 15.9 21s ago
MiniMax M2.7 Pro 64.2 78.8 1.0s 99% 38.9 3m ago
Nemotron 3 Ultra Free 56.1 56.0 2.0s 91% 38.3 43m ago
Kimi K2.6 Pro 52.3 72.4 2.7s 100% 45.1 6m ago
Nemotron 3 Ultra Pro 51.7 33.9 2.7s 89% 38.3 3m ago
GLM 5.1 Pro 50.5 38.9 1.1s 100% 41 7m ago

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

Frequently asked questions

What is the fastest Ollama Cloud model right now?

As of the last build, Nemotron 3 Nano 30B is the fastest Ollama Cloud model on this leaderboard at 371.2 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.