Release 0.37.8

Released 2021-08-03.

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

Release 57441 · Release 57442 · Release 55109

Back to the changelog