gpt-oss-120b

Model overview

Text-only open-weight reasoning and agent model with 117B total and 5.1B active parameters, designed to run on a single 80GB GPU.

general LLMreasoningcodingagentlocal inferencecode generationsoftware engineering agent

Representative version

Company	OpenAI
Release	2025-08-05
Release type	Model release
Parameters	117B / 5.1B active
Architecture	Mixture-of-experts transformer
Context	131,072
Input modalities	text
Output modalities	text
Tool use / function calling	Confirmed: Tool use, Function calling
Structured output / reasoning	Confirmed: Structured output, Reasoning
Vision / audio / video	Not shown: Vision, Audio, Video
Open weights / License	Weights available / Apache-2.0 (license source pending verification) openai-gpt-oss-120b-current · License source pending verification
Vendor API	Not hosted on OpenAI API or ChatGPT gpt-oss Model Card \| OpenAIopenai-gpt-oss-120b-current · 2025-08-05
OpenRouter	OpenRouter listed
Pricing	OpenRouter · 2026-07-15: $0.037 / $0.17 input / output, per 1M tokens
Benchmark	No reliable public source
Public signals	1 community discussion / 1 activity snapshot
Use cases	local reasoning / coding agents / customizable self-hosted workflows / local inference
Claim evidence	Release eventgpt-oss Model Card \| OpenAIgpt-oss-120b · 2025-08-05 Model specs / weightsIntroducing gpt-oss \| OpenAIgpt-oss-120b · gpt-oss-120b

OpenRouter

Pricing / context / availability

OpenRouter	OpenRouter listed
OpenRouter listing	openai/gpt-oss-120b
Mapped version	gpt-oss-120b openai-gpt-oss-120b-current
Pricing	OpenRouter · 2026-07-15: $0.037 / $0.17 input / output, per 1M tokens
Context	131072
Speed/Latency	No reliable public source
Snapshot time	2026-07-15T20:11:41.551Z
Source	View source

Public activity counters

Platform	Entry	Downloads	Likes	Spaces	Time	Source
huggingface	openai/gpt-oss-120b	4,450,991	4,979	100	2026-07-15T20:11:41.551Z	View source

Evidence gaps

No reliable benchmark sourceNo linkable real reviewSpeed/latency not disclosed