Update Agent
Update one or more fields on an existing agent. Only fields you include are changed.
Endpoint
PATCH /agents2/{id}
This page uses Mintlify OpenAPI rendering, including the API tester/runner, request schema, and response schema from the spec.
Updating Pipeline And Speech Settings
Single prompt agents can use either the standardCascaded pipeline or the Speech_To_Speech pipeline. Changing pipeline_type changes which controls apply to the agent version:
| Field | Applies to | Behavior |
|---|---|---|
pipeline_type | Single prompt agents | Cascaded keeps separate transcriber, LLM, and voice controls. Speech_To_Speech uses a realtime speech model. |
speech.transcriber_id | Cascaded agents | Selects the primary STT provider. |
speech.fallback_stt_enabled | Cascaded agents | Enables backup STT. Set speech.stt_fallback_transcriber_id when this is true. |
speech.audio_cache_enabled | Cascaded agents | Enables cached TTS for repeated phrases. Speech to Speech agents ignore this setting. |
Request Example
Request
Response
Authorizations
Pass your API key as a Bearer token in the Authorization header.
Path Parameters
Body
Any subset of agent fields. Unspecified fields are left unchanged.
Set true to publish this agent.
true
Set true to update an existing published agent deployment.
true
Pipeline subtype for single prompt agents. Defaults to Cascaded when omitted.
Cascaded, Speech_To_Speech "Speech_To_Speech"
Response
Success.
"agt_abc123"
"org_xyz"
null
1
false
"2024-03-01T10:00:00.000Z"
"2024-03-01T10:00:00.000Z"
null
Pipeline subtype for single prompt agents. Cascaded uses separate STT, LLM, and TTS components. Speech_To_Speech uses a realtime speech model.
Cascaded, Speech_To_Speech "Cascaded"