> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.jambonz.org/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.jambonz.org/_mcp/server.

#### Paxa Labs

jambonz Commercial (next release)

This entry adds Paxa Labs as a text-to-speech and speech-to-text vendor, for Thai, English and
Mandarin Chinese.

#### New Features & Improvements

1. **Paxa Labs text-to-speech** — `paxa` is available as a TTS vendor, with streaming TTS and
   speech credentials in the portal and API. Each `paxa-tts-flash-v2` voice speaks Thai,
   English and Mandarin Chinese, including in mixed sentences, and voices designed in your
   Paxa account (`voice_…`) work too. Options are `speed` and `language`, direction tags
   such as `[whispering]` are read from the text, and word-level timestamps make alignment
   events available. See [Paxa Labs](/guides/features/tts-vendor-settings/paxa).
2. **Paxa Labs speech-to-text** — the same `paxa` credential serves realtime speech
   recognition with the `paxa-stt-lite-realtime-v1-preview` model, a research preview. Paxa
   detects turns itself and returns one final transcript per turn. `hints` are sent as
   Paxa's vocabulary (up to 50 terms), and `paxaOptions` sets `style`, `convention` and
   word `timestamps`. See [recognizer](/verbs/verbs/recognizer).