Speech-to-Speech: Hydra V1.1 released, ?model=hydra deprecated
Speech-to-Speech: Hydra V1.1 released, ?model=hydra deprecated
Hydra V1.1 is now the default speech-to-speech model on the Waves realtime endpoint. New integrations should connect with ?model=hydra-v1.1:
The session protocol (event catalog, session.configure shape, tool calling, interruption handling) is identical to the original hydra, so switching is a one-parameter change on the query string.
?model=hydra is deprecated. Existing sessions still open, but the server emits a warning frame with code: "model_deprecated" immediately before session.created. Migrate to ?model=hydra-v1.1. See Deprecation Notices.
New voice roster. Ten voice IDs: zoe, maya, elena, ivy, grace, alex, aria, leo, sam, kai. The V1.1 roster does not overlap with the original hydra roster; when you switch versions, also pick a voice from the V1.1 list. Original hydra voices (wren, sloane, marlowe, reed, knox, tate) are unchanged.
Full voice + version reference on the Hydra model card. Overview + docs pages (overview, WebSocket connection, Audio I/O, Managing sessions, Tool calling) point at ?model=hydra-v1.1 in every URL example and use a V1.1 voice in the sample payloads; the voice field description in managing sessions lists both rosters and the model card remains the single source of truth for per-version voice tables.
Pulse STT realtime overview: VAD Events, Endpointing, Age & Gender, and Emotion Detection cards
Pulse STT realtime overview: VAD Events, Endpointing, Age & Gender, and Emotion Detection cards
The Realtime features overview now surfaces four capabilities that had dedicated feature pages but were missing from the overview:
- VAD Events. Acoustic
speech_started/speech_endedframes interleaved with the transcription stream. - Endpointing. Tune the end-of-utterance silence timeout with
eou_timeout_ms. - Age & Gender Detection. Optional per-speaker inference alongside the transcript.
- Emotion Detection. Optional per-utterance classification alongside the transcript.