Voice Agents audio-in / audio-out WebSocket. The connection is established with the target agent identifier; the first client message must be a `setup` frame carrying the API key and the desired audio configuration. Server sends `audioStream` chunks, voice activity events, `newAudioStream` markers, and `error` messages.
First client message on a Voice Agents connection. Carries the API key and the desired audio in/out configuration.
Content-Type: application/json
AgentAudioIn
Streams base64-encoded user audio into the agent.
Content-Type: application/json
AgentAudioStream
Base64-encoded chunk of the agent's spoken response.
Content-Type: application/json
AgentNewAudioStream
Indicates the start of a new agent response stream. Clients should clear their playback buffer and start playing the new stream.
Content-Type: application/json
AgentVoiceActivityStart
Server detected the user started speaking.
Content-Type: application/json
AgentVoiceActivityEnd
Server detected the user stopped speaking.
Content-Type: application/json
AgentError
Error message emitted by the Voice Agents server.
Content-Type: application/json
About AsyncAPI
The AsyncAPI specification describes event-driven APIs the way OpenAPI describes request/response APIs. A channel is the named pipe — a webhook URL, a Kafka topic, a WebSocket route, an MQTT subject — that producers and consumers publish or subscribe to. Each channel carries one or more messages with structured payloads, and an operation declares whether a given party sends or receives on that channel.
Calls without X-API-Key need no key. Calls with it need a paid plan — get a key.
Get an API key
Free tier, no form to fill in. Signing in shares your email address with us — we
store it to create your key and to recognise you if you sign in with another
provider. See our Privacy Policy and
Terms.