About
gpt-oss is OpenAI’s family of open-weight language models, introduced on 5 August 2025. It comprises two models: gpt-oss-120b (117B parameters, 5.1B active) and gpt-oss-20b (21B parameters, 3.6B active). Both are mixture-of-experts reasoning models released under the Apache 2.0 license with full access to the weights, configurable reasoning effort and a visible chain of thought. The models support tool use (web browsing, Python code execution), function calling and structured outputs; they require the harmony response format.
