OpenCode adds stealth coding model Cat Jam Free, free on Zen and Go with no quota used
OpenCode just added Cat Jam Free, a stealth reasoning model built for coding and agent work. On Zen, input, output, and cached reads are all free, and Go subscribers can use it without touching their usage limits. It supports a context window of about 262K tokens and tool calls, and it is free for a limited time with no end date announced yet.
The briefing
OpenCode just added a stealth model called Cat Jam Free (model id cat-jam-free), aimed at coding and agentic workflows. You can call it for free on both OpenCode Zen and OpenCode Go, and no end date has been announced yet.
What you can use
On Zen, input, output, and cached reads for Cat Jam Free are all billed at $0. If you subscribe to Go, this model does not count against your Go usage, and the 5-hour, weekly, and monthly limits all treat it as unlimited.
The provider behind Cat Jam Free has not been named. During the free period the OpenCode team is collecting feedback to improve the model, and your request data may be used to improve it.
Specs
Cat Jam Free is a text-only reasoning model with a context window of about 262K tokens and up to about 131K output tokens. It supports tool calls and structured JSON output, and you can choose a reasoning effort of low, medium, high, xhigh, or max.
How to use it
- Open OpenCode, sign in to Zen, and switch to Cat Jam Free with
/models, or set the model toopencode/cat-jam-freein your config. - If you already subscribe to Go, set the model to
opencode-go/cat-jam-free. - For API calls, use the Zen endpoint
https://opencode.ai/inference/v1/chat/completionswith the model namecat-jam-free.
More details: OpenCode Zen docs · OpenCode Go docs.