---
title: 'Create Call Agent'
api: 'POST /v1/api/agents'
description: 'Create a new agent for a workspace.'
---

> **For AI agents:** the complete documentation index is at [llms.txt](https://docs.trillet.ai/llms.txt). Append `.md` to any page URL for its markdown version.

## Headers
<ParamField header="x-workspace-id" type="string" required>
Workspace identifier for the API.
</ParamField>

## Request Body

<ParamField body="name" type="string" required>
Name of the agent being created.
</ParamField>

<ParamField body="ttsModel" type="object" required>
Text-to-Speech (TTS) configuration for the agent.

<Expandable title="properties">

<ParamField body="provider" type="string" required>
TTS provider. Must be one of the following: openai, rime, elevenlabs, 11labs_byo, google.
</ParamField>

<ParamField body="voiceId" type="string" required>
The voice ID used for generating speech. e.g for Trillet Voices,
"mistv2_Wildflower" - Alt Female, Smart
"mistv2_Brook" - Male, Friendly
"mistv2_Violet" - Female, Professional
</ParamField>


<ParamField body="language" type="string" default='en' required>
Language code (e.g., en, fr).
</ParamField>

<ParamField body="elevenLabsApiKey" type="string">
ElevenLabs API key (required only when provider is "11labs_byo"). This will be encrypted and stored securely.
</ParamField>

</Expandable>
</ParamField>

<ParamField body="smartCallback" type="boolean" default="false">
Enable smart callback functionality for the agent.
</ParamField>

<ParamField body="maxAttemptsForSmartCallback" type="number" default="5">
Maximum number of callback attempts when smart callback is enabled.
</ParamField>

<ParamField body="gapBetweenEachAttemptForSmartCallback" type="array">
Array of time intervals between each callback attempt. Each item contains days and hours.

<Expandable title="Array Item Properties">

<ParamField body="days" type="number" default="1" min="0" max="30">
Number of days to wait before the next attempt.
</ParamField>

<ParamField body="hours" type="number" default="0" min="0" max="23">
Number of hours to wait before the next attempt.
</ParamField>

</Expandable>
</ParamField>

<ParamField body="phoneNumberIds" type="array">
Array of phone number IDs to associate with the agent.
</ParamField>

<ParamField body="llmModel" type="string" default='platform-llm'>
Language model the agent uses to generate responses.

Defaults to `platform-llm`, the managed platform model. It is tuned for voice agents and kept current without any change to your configuration, and it is what agents created in the dashboard start on.

To pin a specific model instead, pass its id, for example `gpt-4.1` or `gemini-2.5-flash`. The models on offer change over time, so call [List Available Models](https://docs.trillet.ai/v1/api-reference/endpoints/agents/call/list-models) for the current set rather than hardcoding one.
</ParamField>

<ParamField body="settings" type="object">
Additional settings for the agent.

<Expandable title="properties">

<ParamField body="speed" type="number" default="0.7">
Value between 0 and 1 that determines how fast the TTS narrates text.
</ParamField>

<ParamField body="volume" type="number" default="1">
Value between 0 and 1 that adjusts the TTS volume.
</ParamField>

<ParamField body="temperature" type="number" default="0.7">
Value between 0 and 1 for controlling the creativity of the LLM.
</ParamField>

</Expandable>
</ParamField>

<ParamField body="variables" type="object">
Custom variables to store for the agent.
</ParamField>

## Response Fields

<ResponseField name="_id" type="string">
Unique identifier for the newly created agent.
</ResponseField>

<ResponseField name="workspaceId" type="string">
Workspace ID associated with the agent.
</ResponseField>

<ResponseField name="userId" type="string">
User ID of the agent creator.
</ResponseField>

<ResponseField name="name" type="string">
Name of the agent.
</ResponseField>

<ResponseField name="ttsModel" type="object">
Text-to-Speech configuration for the agent.

<ResponseField name="provider" type="string">
TTS provider used by the agent.
</ResponseField>

<ResponseField name="voiceId" default="mistv2_Wildflower" type="string">
Voice ID used for generating speech. (e.g., mistv2_Brook, mistv2_Violet, mistv2_Wildflower)
</ResponseField>


<ResponseField name="language" type="string">
Language code of the TTS voice.
</ResponseField>

<ResponseField name="elevenLabsApiKey" type="string">
Encrypted ElevenLabs API key (only present if provider is "11labs_byo").
</ResponseField>
</ResponseField>

<ResponseField name="llmModel" type="string">
Language Model used by the agent to generate responses.
</ResponseField>

<ResponseField name="smartCallback" type="boolean">
Whether smart callback is enabled for the agent.
</ResponseField>

<ResponseField name="maxAttemptsForSmartCallback" type="number">
Maximum number of callback attempts.
</ResponseField>

<ResponseField name="gapBetweenEachAttemptForSmartCallback" type="array">
Array of time intervals between callback attempts.
</ResponseField>

<ResponseField name="phoneNumberIds" type="array">
Array of phone number IDs associated with the agent.
</ResponseField>

<ResponseField name="activeSessions" type="array">
Array of active sessions for the agent.
</ResponseField>

<ResponseField name="settings" type="object">
Additional settings configured for the agent.
</ResponseField>

<ResponseField name="status" type="string">
Current status of the agent (pending, ready, or error).
</ResponseField>

<ResponseField name="currentNode" type="string">
Current node or state within the agent's flow.
</ResponseField>

<ResponseField name="variables" type="object">
Custom variables stored for the agent.
</ResponseField>

<ResponseField name="createdAt" type="string">
Timestamp when the agent was created.
</ResponseField>

<ResponseField name="updatedAt" type="string">
Timestamp when the agent was last updated.
</ResponseField>

<ResponseExample>
```json
{
  "name": "Customer Support Agent",
  "ttsModel": {
    "provider": "rime",
    "voiceId": "mistv2_Wildflower",
    "name": "Sophie",
    "language": "en"
  },
  "llmModel": "platform-llm",
  "type": "voice",
  "smartCallback": true,
  "maxAttemptsForSmartCallback": 3,
  "gapBetweenEachAttemptForSmartCallback": [
    { "days": 1, "hours": 2 },
    { "days": 2, "hours": 0 },
    { "days": 3, "hours": 6 }
  ],
  "settings": {
    "speed": 0.8,
    "volume": 1,
    "temperature": 0.7
  }
}
```
</ResponseExample>