Release 0.25.8

Released 2023-03-02.

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

Release 21297 · Release 21298 · Release 71892

Back to the changelog