Skip to main content
GET
Get all voice models associated with a specific voice
Legacy endpoint - deprecated. Unversioned endpoints are scheduled for removal on July 31, 2026. A public v1 replacement is not currently listed in the v1 reference. This endpoint remains available for existing integrations until then.
Returns the text-to-speech model variants available for a given voice. Each entry corresponds to a deployable model (for example eleven_multilingual_v2, eleven_turbo_v2) and includes the upstream provider. Different variants of the same voice trade off latency, naturalness, and cost. Use this endpoint to populate the model selector in your agent build UI and to confirm a saved model_id is still supported before publishing an agent change.

When to use this

  • Voice configuration UIs: let the prompt author choose between latency-optimized and quality-optimized variants of the same voice.
  • Latency tuning: switch a real-time agent to a “turbo” model when first-word latency matters more than expressiveness.
  • Cost optimization: switch a high-volume outreach use case to a cheaper variant if naturalness is acceptable.
  • Migration safety checks: confirm the saved model_id is still in the response before deploying an agent update.

Common trade-offs

Across providers, voice models tend to cluster into three loose tiers:
  • Turbo / low latency: fastest first-word time. Best for live conversational agents where caller perception of speed matters most.
  • Multilingual / balanced: moderate latency with broad language support. The default for most production deployments.
  • High fidelity / premium: best naturalness and prosody. Typically reserved for branded greetings or scripted spoken content rather than free-flowing live conversation.

Path parameters

Errors

  • 404 Not Found is returned when the voice does not exist or has been removed.

Authorizations

Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Path Parameters

id
string
required

Response

Returns all voice models linked to the specified voice