FishAudioSTT
Realtime speech-to-text with Fish Audio’s transcribe-stream-1, using your own Fish Audio key. Server-side turn detection tells the user apart from noise and backchannels. Authenticates with FISH_API_KEY from your environment, or an explicitapi_key.
Fields
str
default:"UNSET"
Realtime transcription model. “transcribe-stream-1” (Preview) is the only one. Defaults to ‘transcribe-stream-1’.
Literal['auto', 'vad', 'manual']
default:"UNSET"
How a turn ends: “auto” lets Fish Audio judge when the user has finished, “vad” ends it on fixed silence, “manual” only on flush. Defaults to ‘auto’.
int | None
default:"UNSET"
Silence that ends a turn, in milliseconds. Leave unset for the mode’s default: 5000 in auto (a safety net), 700 in vad, 60000 in manual. Defaults to None (not set).
list[str] | None
default:"UNSET"
Up to 100 names, products and jargon terms (32 characters each) to recognize, e.g. [“Fish Audio”]. Defaults to None (not set).
bool
default:"UNSET"
Start the reply on Fish Audio’s eager end-of-turn signal, about half a second before the turn is committed. Auto mode only; set False to wait for the committed transcript. Defaults to False.
int | None
default:"UNSET"
Override the detected input sample rate. Leave unset in normal use. Defaults to None (not set).
int | None
default:"UNSET"
Override the detected input channel count. Leave unset in normal use. Defaults to None (not set).
str
default:"UNSET"
WebSocket endpoint of Fish Audio realtime speech-to-text. Change only for a proxy. Defaults to ‘wss://api.fish.audio/v1/asr/realtime’.
bool
default:"UNSET"
Set to True to also send live partial transcripts to the client. Defaults to False.