Release 0.12.2

Released 2011-10-19.

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

Release 88335 · Release 88336 · Release 38025

Back to the changelog