Release 0.3.5

Released 2019-07-09.

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

Release 87630 · Release 87631 · Release 73188

Back to the changelog