Skip to content

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

Before you go

Comparing more than Ollama?

Ollama TPS measures Ollama Cloud. TokenDyno is our sister board — the same continuous measurement across Ollama Cloud, OpenCode Zen and OpenCode Go, so you can compare providers side by side rather than one at a time.

See the multi-provider board

Leaderboard Compare Pricing Limits vs OpenRouter Methodology About

Get Ollama Cloud →

Compare models

Pick up to 6 models and compare their tokens-per-second history side by side. The chart link is shareable — selections are in the URL.

Models

0 / 6

DeepSeek V4 Flash 0731 215.0 t/s
DeepSeek V4 Pro 0813 141.1 t/s
Gemma4 31B 362.0 t/s
Gemma4 31B 296.8 t/s
GLM 5.1 85.6 t/s
GLM 5.2 27.1 t/s
GLM 5.3 93.9 t/s
GLM 5.3 Flash 64.8 t/s
GPT-OSS 120B 367.9 t/s
GPT-OSS 120B 158.1 t/s
GPT-OSS 20B 83.5 t/s
GPT-OSS 20B 106.1 t/s
Kimi K2.6 30.1 t/s
Kimi K2.7 Code 104.6 t/s
Kimi K3 108.5 t/s
MiniMax M2.7 30.8 t/s
MiniMax M3 282.7 t/s
MiniMax M3 82.2 t/s
Mistral Large 3 675B 69.2 t/s
Nemotron 3 Nano 30B 99.8 t/s
Nemotron 3 Nano 30B 118.4 t/s
Nemotron 3 Super 68.3 t/s
Nemotron 3 Super 103.8 t/s
Nemotron 3 Ultra 71.1 t/s
Nemotron 3 Ultra 99.8 t/s
Qwen3.5 397B 104.7 t/s

No models match that filter.

Want provider vs provider, not just model vs model?

The same benchmark runs across Ollama Cloud, OpenCode Zen and OpenCode Go on TokenDyno — compare providers side by side.

Open TokenDyno compare →

Benchmarked every ~10 minutes. Models Pricing Limits Privacy TokenDyno — multi-provider board Support this project ♥ Get Ollama Cloud © 2026 Ollama TPS