GPT-OSS 120B
Use it free in free-pi
~1.6x usage
npx free-pi-cli
/model gpt-oss-120b-speedAbout GPT-OSS 120B (Speed)
gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts language model from OpenAI, designed for high-reasoning, agentic, and general-purpose production use. It activates 5.1B parameters per forward pass, which keeps it quick. In free-pi it is served on a fast speed lane, so replies come back sooner than from most models here. It became available in August 2025.
The model takes text in and produces text. Its context window is about 131,000 tokens, and a response can be up to about 65,500 tokens. It supports reasoning with an effort setting, tool calling, and structured outputs. With tool calling, the free-pi agent can read and edit files, run commands, and use free-pi's built-in web search and documentation lookup tools.
It scores 30.4 on the Artificial Analysis Coding Index.
In free-pi it is free to use with no API key and no card. It counts as about 1.6x usage, so it stretches your daily allowance further than the default.
When to pick it
For fast responses on small, well-defined tasks, such as quick questions about a file, small edits, or short scripts, pick GPT-OSS 120B (Speed). It also stretches your allowance further than the default, and its open weights make it a good choice if you prefer an open model.
For large codebases and longer multi-step agent work, pick the default DeepSeek V4 Flash, which has a context window about eight times larger. For the strongest results on a hard task, pick a premium model such as Claude Sonnet 5.5.