Release 0.4.0

Released 2026-01-05.

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

Release 48037 · Release 48038 · Release 21896

Back to the changelog