ConnectRequest

Sinch Voice API: Choosing between v1 and v2

For all new projects, please use Voice API v2. This is especially important for projects involving:

AI capabilities like Streams and Voice Relay
Voice campaigns like Batch calling

You should only use the older Voice API v1 if one of the following is a strict requirement:

The project must be in General Availability (GA).
The project requires In-App voice or Conferencing (until these features are available in v2).

ConnectRequest

Payload Schema

command (string, required) Connect message
version (integer, required) Protocol version
callId (string, required) Call unique identifier
applicationKey (string, required) Service unique identifier
codec (string, required) Audio codec used for encoding binary audio frames transmitted over the WebSocket connection. Audio frames are 16-bit signed linear PCM, mono (single channel), with little-endian byte order.
sampleRate (integer, required) Audio sample rate in Hertz (Hz). The value is negotiated at call setup via the SVAML connectStream command. Higher sample rates provide better audio quality but increase bandwidth and processing requirements.
8000 Hz: Narrowband telephony quality (standard PSTN)
16000 Hz: Wideband quality, suitable for most voice applications
24000 Hz: High-quality voice
44100 Hz: CD-quality audio
48000 Hz: Professional audio standard
96000 Hz: Ultra high-quality audio

For calls traversing the PSTN, audio is typically sampled at 8 kHz; using a higher value will not improve perceived quality. Enum: 8000, 16000, 24000, 44100, 48000, 96000

callHeaders (object, required) Arbitrary key/value pairs (header name -> header value).