← LLM

Baseten

gpt-oss-120b

reasoning off (gateway default)

Measured 2026-08-17

* Conditional

Silent on every tool call; matches its Cerebras twin within 3pp on four axes.

Score

73

Task done

83%

66–93% 95% CI · n=30

Fabrication

14%

Dead-air

54%

Tool silence

100%

Stalled

25%

TTFT p50

393ms

393–536 p50–p90

Cost / 1M tok

$0.50

Share this result

Download card (PNG) Post on X

Embed the live badge

Speko llm rank
Markdown
[![Speko llm rank](https://benchmarks.speko.ai/badge/llm/baseten-gpt-oss-120b.svg)](https://benchmarks.speko.ai/llm/baseten-gpt-oss-120b)
HTML
<a href="https://benchmarks.speko.ai/llm/baseten-gpt-oss-120b"><img src="https://benchmarks.speko.ai/badge/llm/baseten-gpt-oss-120b.svg" alt="Speko llm rank"></a>
URL
https://benchmarks.speko.ai/llm/baseten-gpt-oss-120b