gpt-oss:120b tokens per second
Live benchmark data for gpt-oss:120b on ollama. Speed, latency, and reliability — benchmarked every ~10 minutes.
- Latest TPS
- 141.3 tok/s
- 24h avg TPS
- 142.3 tok/s
- TTFT
- 438ms
- 24h reliability
- 100%
- Last tested
- 3m ago
Tokens per second over time
Time to first token over time
Speed is measured as output tokens ÷ generation time. Each benchmark fires one streaming chat completion with a fixed ~300-token prompt, every ~10 minutes. Full methodology →