vtrix models

vtrix models list

List available models with optional type filtering and keyword search.

vtrix models list                          # List all models
vtrix models list --type video             # Filter by type: video / image / audio
vtrix models list --keywords kirin         # Search by keyword
vtrix models list --output id              # Output model IDs only (good for scripts)
vtrix models list --page 2 --page-size 20  # Pagination

Flag	Description
--type	Filter by model type: video / image / audio
--keywords	Keyword search
--output	id outputs model ID list only / json full JSON
--page	Page number, default 1
--page-size	Results per page, default 20

Example output:

vtrix models list --type video --keywords spark

Models (2 results)

  spark_dance_v2_0       Seedance 2.0        video   text/image/video/audio
  spark_dance_v2_0_fast  Seedance 2.0 Fast   video   text/image/video/audio

Page 1 / 1  ·  Use --output json for full details

vtrix models list --type video --output id

spark_dance_v2_0
spark_dance_v2_0_fast
kirin_v2_6_t2v
kirin_v2_6_i2v
kirin_v3_t2v
...

---

vtrix models spec

Fetch the parameter spec for a given model from the server in real time, ensuring agents and developers always have the latest, accurate parameter definitions.

vtrix models spec <model_id>               # Formatted text output (good for injecting into agent context)
vtrix models spec <model_id> --output json # Raw JSON (good for programmatic parsing)

---

Demo: Spark Dance 2.0

Using spark_dance_v2_0 (Seedance 2.0, the flagship video model) as an example.

View parameter spec

vtrix models spec spark_dance_v2_0

Model: spark_dance_v2_0
Name:  Seedance 2.0
Type:  video

Parameters:
  content            array     required   Multimodal input array
                                          Supported types: text | image_url | video_url | audio_url

  resolution         string    optional   Output resolution
                                          Values: 480p | 720p | 1080p
                                          Default: 720p

  ratio              string    optional   Aspect ratio
                                          Values: 16:9 | 9:16 | 1:1 | 4:3 | 3:4 | 21:9 | adaptive
                                          Default: adaptive

  duration           integer   optional   Video duration in seconds, range 4–15, -1 for auto
                                          Default: 5

  generate_audio     boolean   optional   Whether to generate synchronized audio
                                          Default: true

  seed               integer   optional   Random seed, -1 for random
                                          Default: -1

  return_last_frame  boolean   optional   Whether to return the last frame as an image
                                          Default: false

  service_tier       string    optional   Inference mode
                                          Values: default | flex (flex = async inference, 50% off)
                                          Default: default

  camera_fixed       boolean   optional   Fix the camera angle
                                          Default: false

  callback_url       string    optional   Callback URL on task completion

  safety_identifier  string    optional   End-user identifier (≤ 64 characters)

Get JSON format (for programmatic parsing)

vtrix models spec spark_dance_v2_0 --output json

{
  "model_id": "spark_dance_v2_0",
  "name": "Seedance 2.0",
  "type": "video",
  "parameters": {
    "content": { "type": "array", "required": true },
    "resolution": { "type": "string", "required": false, "default": "720p", "enum": ["480p","720p","1080p"] },
    "ratio": { "type": "string", "required": false, "default": "adaptive" },
    "duration": { "type": "integer", "required": false, "default": 5, "min": 4, "max": 15 },
    "generate_audio": { "type": "boolean", "required": false, "default": true },
    "seed": { "type": "integer", "required": false, "default": -1 },
    "service_tier": { "type": "string", "required": false, "default": "default", "enum": ["default","flex"] }
  }
}

Generate a video with this model

After checking the spec, submit a task directly with vtrix run:

# Text to video, 16:9, 1080p, 10 seconds
vtrix run spark_dance_v2_0 \
  --param prompt="cinematic drone shot over misty mountain peaks at sunrise" \
  --param resolution=1080p \
  --param ratio=16:9 \
  --param duration=10 \
  --output url

Submitting task...
✓ Task created: task_abc123

Waiting for result...
████████████████████░░░░  82%  ~18s remaining

✓ Done
  https://cdn.vtrix.ai/outputs/task_abc123/output.mp4