Skip to content

v0.0.1 beta-2

Pre-release
Pre-release

Choose a tag to compare

@mistqkw mistqkw released this 26 Oct 06:20
· 5 commits to main since this release

v0.0.1 beta-2

  • When the built-in free Zen models hit their limit, the app switches to Pollinations (GPT-OSS 20B, no account) and resends the last message. In that mode messages go to pollinations.ai.
  • Fallback options: GitHub Copilot (GPT-4.1, GPT-4o, GPT-4o mini) or your own provider key in Settings → Providers → Free models.
  • "Free limit reached" card: the Stop button is now red and full width, the other buttons wrap to a new line.
  • Memory: the agent now really writes facts to the file. Chat fonts follow the chosen interface.

The Zen limit is counted by the OpenCode server and happens in the official client too. This is a beta and there are bugs.