Changelog

August 9, 2026

• Added incremental KittenTTS audio streaming in every output format.
• Added OpenAI-style stream_format: audio and stream_format: sse responses.
• Added runnable Python and Node.js examples for binary and SSE speech streams.

August 8, 2026

• Documented the production api.kittenml.com hostname and the three primary inference routes.
• Added upload STT JSON, verbose JSON, text, SRT, VTT, and finite SSE output.
• Added realtime STT on WSS /v1/stt/realtime with partial and final events.
• Added free KittenTTS 0.8 Nano, Micro, and Mini generation.
• Documented organization concurrency limits, error behavior, and the tested OpenAI SDK compatibility surface.