Release 0.3.6

Released 2026-01-29.

• Fixed token counting for long prompts
• Faster model list caching
• Added per-model rate limits

Release 21031 · Release 21032 · Release 49076

Back to the changelog