Release 0.32.8

Released 2015-09-04.

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

Release 45981 · Release 45982 · Release 32124

Back to the changelog