Models

Choose a model based on the experience you want to build.

A-series — for realtime avatar experiences and avatar video generation.
W-series — for continuous realtime worlds and reference-guided world video generation.

Pass one of the supported model IDs below in the model field. Refer to the linked API guide for endpoint-specific requirements. To discover which models are enabled for your API key, call GET /v1/models.

Model overview

Model	What it does	Docs
`vivix-a1-stream`	Realtime avatar sessions	Streaming Avatar
`vivix-a1`	Avatar videos driven by script, audio and images	Avatar Video Generation
`vivix-w1-stream` Private beta	Continuous realtime world sessions steered by open-ended Events	Streaming World
`vivix-w1`	Reference-driven world video generation. Also available as `vivix-w1-fast` and `vivix-w1-ultrafast` for faster generation with the same video quality.	World Video Generation

Private beta. vivix-w1-stream is not generally available yet. It only appears on GET /v1/models for approved workspaces, and requests using it are rejected otherwise. See Streaming World.

List Models

Returns the models enabled for your API key. Use this to discover which name values you can pass as model on your requests, across Streaming Avatar, Avatar Video Generation, Streaming World, and World Video Generation. The list is scoped to your workspace: you only see models that are currently enabled for your key. Each entry is one way you are allowed to call a model, so the same name can appear more than once with a different tag.

Response envelope. REST responses are wrapped in {code, message, data}: code = 0 means success. Field descriptions below refer to fields inside data.

GET /v1/models

Response Fields

array

Models available to the current API key. An empty list means none are enabled for this key.

Show properties

string

Public model ID. Pass this as the model field on later requests.

optional string

An available version tag or alias. Pass it as name:tag, for example vivix-a1-stream:1.0.7 or vivix-a1-stream:latest, when that value appears in your response.An entry without tag allows the bare model name. An entry with tag: "latest" allows the explicit name:latest form. These are separate entries; latest is not automatically omitted.

Error Cases

Condition	API error code
The API key is missing or invalid.	`10001 missing api key` or `10003 invalid api key`
The request exceeded the API key rate limit (HTTP 429; 60 requests/min per key by default; the response includes a Retry-After header).	`10008 API rate limit exceeded`

Use a model name and tag

Use the model name alone when your response includes an entry without tag. For an entry with tag, join the name and tag with a colon:

{ "model": "vivix-a1-stream:1.0.7" }

The bare name, name:latest, and name:1.0.7 have separate access permissions. Access to one does not grant access to the others. Use a form returned for your API key. A concrete version tag selects that version. An alias such as latest can resolve to a different version over time, so it does not freeze a version. The bare name also follows the currently assigned version.

Request

curl https://api.vivix.ai/v1/models \
  -H "Authorization: Bearer $VIVIX_API_KEY"

Response

{
  "code": 0,
  "message": "success",
  "data": {
    "models": [
      {
        "name": "vivix-a1"
      },
      {
        "name": "vivix-a1-stream"
      },
      {
        "name": "vivix-w1"
      }
    ]
  }
}

The exact set depends on your API key, so treat the response above as a shape example, not a fixed list. Entries for a concrete version or an alias include tag.

Overview

A1: Streaming Avatar

W1: Streaming World

Blazing Fast Video Generation

Model overview

List Models

Response Fields

Error Cases

Use a model name and tag