Mode
Streaming workloads sit on a spectrum between two competing goals: returning transcripts as fast as possible, and returning the most accurate transcripts possible. To make this tradeoff explicit, every streaming model supports a mode connection parameter you can set when opening a streaming session. On AssemblyAI-STT-3.7-Realtime, turn detection is tuned for transcript segmentation rather than voice-agent turn-taking (see Turn detection on AssemblyAI-STT-3.7-Realtime).
Set the
mode connection parameter when you open the WebSocket.
- Python
- Python SDK
- Javascript
- JavaScript SDK
Language selection
By default, streaming is multilingual and code-switches natively across all supported languages with no configuration. When you know which languages a session will use, pass thelanguage_codes connection parameter to steer the model toward them and improve language accuracy.
- For a known subset, pass the codes you expect (for example,
["en", "es"]). The model still code-switches, but heavily biased to the languages you list. - For a monolingual session, pass a single-element list (for example,
["es"]). - For full multilingual, omit
language_codesto keep native code switching.
language_codes mid-stream.
Turn detection
The model detects the end of a turn using the transcript rather than silence alone. Themode preset sets the defaults, and every turn detection parameter can be overridden on the connection or updated mid-stream to tune endpointing for your use case.
- Raise
min_turn_silencewhen brief pauses end turns too early, for example while a caller dictates a phone number. Raisemax_turn_silencewhen you expect longer pauses within a turn. - Lower
vad_thresholdwhen quiet speech is missed. When background noise causes false interruptions, raisevad_threshold, increaseinterruption_delay, or enable Voice Focus.