Formats
Use these formats across the API.
Timestamps
Timestamps use UTC RFC 3339 /
ISO 8601, with a Z suffix and a
three-digit millisecond fraction:
Phone numbers
Phone numbers use canonical international E.164
with a leading +, for example +972500000000.
WebSocket audio
WebSocket Sessions exchange raw mono audio in binary messages without
a WAV header. The required media_format sets both directions from the Session’s
perspective:
from_session: Session to your WebSocket server.to_session: Your WebSocket server to Session.
Supported profiles:
Select each direction independently. The platform converts Session audio to the
from_session profile and to_session audio to the Session media rate. Message
sizes may vary. PCM16 messages must contain complete 16-bit samples; each mu-law
byte is one complete sample.
PCM16 messages can contain at most 64 KiB. A to_session mu-law message can
contain at most 32 KiB, keeping its PCM16 expansion within that bound.
G.711 mu-law changes representation, not sample rate: mulaw uses exactly 8
kHz. pcm_s16le can use any supported multiple of 8 kHz.
The directions may differ. For example, your server can receive pcm_s16le at
16 kHz in from_session and send mulaw at 8 kHz in to_session. See Create a
WebSocket Session for the schema.
Audio is raw. mulaw does not add Twilio Media Streams JSON envelopes, base64
payloads, or control messages.