InfiniteTalk · VIDEO

Talking Avatars API

Create video with Talking Avatars.

Available through Sorceress API

Talking Avatars model artwork

Using Talking Avatars in your workflow

Sorceress brings InfiniteTalk models into a shared API catalog. Select this model explicitly, check its supported inputs, and keep your API key on your server. Model capabilities differ; the schema and options on this page describe this entry rather than every model from the same brand.

  • Explore motion and scene concepts.
  • Automate video generation with a saved job that your application can track.

Pricing and availability

Cost scales with audio length and resolution (480p or 720p). Charged only on success. Call list tools for current per-second examples.. Successful work is billed in Sorceress credits. Defaults and optional inputs can affect the total.

Resolution

480p (default)122 credits

720p262 credits

Inputs and supported options

imageUrlstring · required

https URL of a portrait / character still (JPEG, PNG, or WebP). Use images[0].url from image_generate, or file_upload publicUrl.

audioUrlstring · required

https URL of the driving voice clip (MP3/WAV/AAC/OGG). Max 15 seconds. Use speech_generate result.audio.url or file_upload.

promptstring · optional

Motion / performance direction (default: a natural talking-head). Required by the model; a default is used if omitted.

resolutionstring · optional

480p (default, cheaper) or 720p.

audioSecondsnumber · optional

Clip length in seconds (1–15). Optional — probed from audioUrl when omitted. Needed if the server cannot read the file.

API request template

Replace placeholders and choose valid values from the input contract.

POST /api/v1/tools/talking_avatar
Authorization: Bearer $SORCERESS_API_KEY
Content-Type: application/json

{
  "imageUrl": "<imageUrl>",
  "audioUrl": "<audioUrl>"
}
Read the endpoint reference →

Completion, results and billing

Save the job identifier returned by asynchronous requests and poll the jobs endpoint until the job succeeds or fails. Do not resubmit a generation just because it is still processing. Chat returns a completed reply and usage details. A multi-step workflow can include separately billed successful steps even if a later step fails. Keep returned asset links with your project, and consult each action for its output format.

Coding agents can discover the current catalog and input requirements from the Sorceress agent instructions. Use a Sorceress key; provider credentials stay on the server.