← Back to models

GPT-OSS

OpenAI (USA) August 5 2025

Parameters

~120B / 5.1B (MoE) and ~21B / 3.6B (MoE)

License

Apache 2.0

Key Features

First OpenAI open-weight release since GPT-2 (2019); both models use MoE architecture with MXFP4 quantization; 120B for reasoning and complex tasks (fits single H100); 21B for lightweight applications and on-device deployment (runs on 16GB); 128K context; adjustable reasoning effort levels; strong on code generation and structured reasoning.

Paper / Source

https://huggingface.co/blog/welcome-openai-gpt-oss