Release 0.18.9

Released 2017-07-03.

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

Release 6814 · Release 6815 · Release 3813

Back to the changelog