Skip to main content

AI Models

Tags: ai, configuration, models

Endpoints​

All​

GET /api/v2/ai-models/all

Description:

Available AI models and capabilities

List all AI models with full metadata.

Returns all model types (chat, realtime, image, stt, tts) with complete metadata including multipliers for a units calculation.

This is the recommended endpoint for clients that need to display comprehensive model information or perform billing calculations.

Each model contains:

  • id: globally unique identifier
  • name: human-readable name
  • provider: provider name
  • type: model type (chat, realtime, image, stt, tts)
  • description: optional description
  • state: "enabled", "disabled", or "deprecated"
  • available: whether the model is currently available
  • recommended: whether this model is recommended
  • replacement: ID of replacement model if deprecated
  • multiplier: credits multiplier for billing - 1 credit === 1 ai unit

Query parameters:

  • provider: Filter by provider
  • include_unavailable: Include unavailable models (default: false)

Authentication: Requires authenticated user

Parameters:

  • provider (ModelProvider)
  • include_unavailable (Boolean)

Response: See AllModelsResponse

Chat​

GET /api/v2/ai-models/chat

Description:

Available AI models and capabilities

List all available chat AI models.

Returns chat models with optional filtering by provider. Each model contains:

  • id: globally unique identifier (e.g., "openai:chat:gpt-4.1")
  • name: human-readable name
  • provider: provider name (openai, google, anthropic, aws)
  • type: "chat"
  • description: optional description
  • state: "enabled", "disabled", or "deprecated"
  • available: whether the model is currently available for use
  • multiplier: credit multiplier for billing

Query parameters:

  • provider: Filter by provider (openai, google, anthropic, aws)

Authentication: Requires authenticated user

Parameters:

  • provider (ModelProvider)

Response: See ChatModelsResponse

Image​

GET /api/v2/ai-models/image

Description:

Available AI models and capabilities

List all available image generation AI models.

Returns image models with optional filtering by provider. Each model contains:

  • id: globally unique identifier (e.g., "openai:image:dall-e-3")
  • name: human-readable name
  • provider: provider name (openai, google)
  • type: "image"
  • description: optional description
  • state: "enabled", "disabled", or "deprecated"
  • available: whether the model is currently available for use
  • multiplier: credit multiplier per image generated

Query parameters:

  • provider: Filter by provider (openai, google)

Authentication: Requires authenticated user

Parameters:

  • provider (ModelProvider)

Response: See ImageModelsResponse

Realtime​

GET /api/v2/ai-models/realtime

Description:

Available AI models and capabilities

List all available realtime AI models.

Returns realtime voice models with optional filtering by provider. Each model contains:

  • id: globally unique identifier (e.g., "openai:realtime:gpt-4o-realtime")
  • name: human-readable name
  • provider: provider name (openai, google, aws)
  • type: "realtime"
  • description: optional description
  • state: "enabled", "disabled", or "deprecated"
  • available: whether the model is currently available for use
  • voices: list of supported native voice model IDs
  • serverVad: whether server-side VAD is supported
  • clientVad: whether client-side VAD is supported
  • multiplier: credit multiplier for billing

Query parameters:

  • provider: Filter by provider (openai, google, aws)

Authentication: Requires authenticated user

Parameters:

  • provider (ModelProvider)

Response: See RealtimeModelsResponse

Stt​

GET /api/v2/ai-models/stt

Description:

Available AI models and capabilities

List all available speech-to-text (STT) AI models.

Returns STT models with optional filtering by provider. Each model contains:

  • id: globally unique identifier (e.g., "deepgram:stt:flux-general-en")
  • name: human-readable name
  • provider: provider name (deepgram, elevenlabs, openai, google)
  • type: "stt"
  • description: optional description
  • state: "enabled", "disabled", or "deprecated"
  • available: whether the model is currently available for use
  • multiplier: credit multiplier per minute of audio processed

Query parameters:

  • provider: Filter by provider (deepgram, elevenlabs, openai, google)

Authentication: Requires authenticated user

Parameters:

  • provider (ModelProvider)

Response: See STTModelsResponse

Tts​

GET /api/v2/ai-models/tts

Description:

Available AI models and capabilities

List all available text-to-speech (TTS) AI models.

Returns TTS models with optional filtering by provider. Each model contains:

  • id: globally unique identifier (e.g., "elevenlabs:tts:eleven_turbo_v2_5")
  • name: human-readable name
  • provider: provider name (elevenlabs, openai, google)
  • type: "tts"
  • description: optional description
  • state: "enabled", "disabled", or "deprecated"
  • available: whether the model is currently available for use
  • multiplier: credit multiplier per 1000 characters

Query parameters:

  • provider: Filter by provider (elevenlabs, openai, google)

Authentication: Requires authenticated user

Parameters:

  • provider (ModelProvider)

Response: See TTSModelsResponse

Voices​

GET /api/v2/ai-models/voices

Description:

Available AI models and capabilities

List available voice models with metadata.

Returns all voice models with their names, descriptions, and provider info. Voice IDs are in the format: provider:voice:name (e.g., elevenlabs:voice:catt)

Query parameters:

  • provider: Filter by provider (openai, google, aws, elevenlabs, deepgram)

Authentication: Requires authenticated user

Parameters:

  • provider (ModelProvider)

Response: See VoiceModelsResponse