Release 0.3.0

Released 2026-02-22.

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

Release 15025 · Release 15026 · Release 84310

Back to the changelog