Skip to main content
GET
List LLM Models With Voices

Endpoint Overview

GET /api/v1/available-llm-models/with-voices Returns the platform’s available LLM models along with the voices supported for each model. Use voices[].id as the voice_id when creating or updating agents.

Response Structure

The response is a JSON object with the following top-level fields:
  • success — boolean
  • message — string
  • data — array of model objects (see Model Object Schema)
  • meta — pagination metadata

Root Fields Explanation

  • success: Whether the API request succeeded.
  • message: Human-readable message for the result.
  • data: Array of model objects. Each model contains a voices array.
  • meta: Pagination details (total, limit, offset).

Model Object Schema

A model object describes an LLM and metadata about available voices.
  • id (string, UUID): Model identifier
  • llmModel (string): Model name
  • description (string): Short description of the model
  • price (number): Pricing metric (units depend on plan)
  • isActive (boolean): Whether the model is active
  • voices (array): List of voices supported by this model

Voice Object Schema

Each voice object inside voices has the following fields:
  • id (string, UUID): Unique voice identifier
  • llmModelId (string, UUID): Parent model ID
  • voiceName (string): Voice display name
  • gender (string): Voice gender label (optional)
  • language (string): Language/locale code (e.g., en-US)
  • recordingUrl (string, URL): Sample audio file URL
  • imgUrl (string, URL): Avatar image URL

Pagination Meta Schema

  • meta.total (integer | string): Total number of records
  • meta.limit (integer): Limit used for the request
  • meta.offset (integer): Offset used for the request

Example Response

Below is a complete example response containing the two models Agni Lite and Agni Premium with multiple voices each.

Field Descriptions Table

Model fields Voice fields

Voice Relationship with Models

  • Each voice belongs to a single model via llmModelId.
  • You can filter or lookup voices by model ID.
  • Use voices[].id as voice_id when creating agents to bind a voice to the agent.

Notes and Validation Rules

  • id, llmModelId must be valid UUIDs.
  • recordingUrl and imgUrl must be valid HTTPS URLs.
  • price is expressed in the platform’s billing units; check plan docs for currency/scaling.
  • isActive: false models/voices should not be offered for new agents.
  • language should use BCP-47 locale codes (e.g., en-US).

If you want the example to include the full lists of every voice name for Agni Lite and Agni Premium (all names you provided), I can expand the voices arrays in the example to include every voice entry with sample recording and image URLs.

Authorizations

X-Api-Key
string
header
required

Query Parameters

limit
integer<int32>

The maximum number of model records to return.

offset
integer<int32>

The number of model records to skip before returning results.

Response

200 - application/json

LLM models with voices fetched successfully

success
boolean
message
string
data
object[]
meta
object