Platform
Checking your session…
← All models

MetaSAM 3 VideoVideo → Video

Segment and track objects through a video from a text prompt — name several objects separated by commas and SAM 3 keeps each one masked across every frame.

SegmentationObject TrackingVideo Input

Authentication

Send your Picsart API key as a bearer token on every request.

Header
Authorization: Bearer <PICSART_API_KEY>

Make a request

One endpoint serves every model: POST to /workflows/v1/models/execute with the model id and its input in the body. The nesting is deliberate — the outer params is the envelope every workflow takes, and the inner one is this model's own input. Those inner names are the same unified ones the SDK takes, so a call translates one-to-one between the two — no per-vendor naming to learn.

cURL
curl -X POST 'https://api.picsart.com/workflows/v1/models/execute' \
  -H 'Authorization: Bearer $PICSART_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "params": {
    "model": "sam-3-video",
    "params": {
      "prompt": "A serene mountain lake at golden hour, ultra detailed",
      "videoUrl": "https://cdn.example.com/input.mp4"
    }
  }
}'

Async (submit & poll)

This model can run longer than the sync limit (~20s), so send it to /workflows/v1/models/submit, which returns a task id, and poll /workflows/v1/models/{taskId}/result until response.status is COMPLETED or FAILED.

cURL
# 1. Submit the job
curl -X POST 'https://api.picsart.com/workflows/v1/models/submit' \
  -H 'Authorization: Bearer $PICSART_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "params": {
    "model": "sam-3-video",
    "params": {
      "prompt": "A serene mountain lake at golden hour, ultra detailed",
      "videoUrl": "https://cdn.example.com/input.mp4"
    }
  }
}'
# → { "status": "success", "response": { "id": "task_abc123" } }

# 2. Poll until response.status is COMPLETED
curl 'https://api.picsart.com/workflows/v1/models/task_abc123/result' \
  -H 'Authorization: Bearer $PICSART_API_KEY'
Result body
{
  "status": "success",
  "response": {
    "id": "task_abc123",
    "status": "COMPLETED",
    "result": {
      "url": "https://cdn.picsart.com/…/result",
      "items": [
        {
          "url": "https://cdn.picsart.com/…/result"
        }
      ],
      "model": "sam-3-video"
    }
  }
}

Parameters

5 parameters, sent inside the inner params object. Identical to the SDK's — one parameter set, whichever transport you use.

ParameterTypeRequiredDefaultDetails
prompt
Prompt
textyes—e.g. "person, cloth"
videoUrl
Source Video
file (video)yes——
detectionThreshold
Detection Threshold
rangeno0.50–1 (step 0.05)
applyMask
Apply Mask
booleannotrue—
outputFormat
Format
enumnomp4mp4, webm

Response

The body arrives in the gateway's status envelope, so the output is at response.result — alongside the task id, its status and the amount charged. Asset models return url plus an items array; text models return text instead.

Response body
{
  "status": "success",
  "response": {
    "id": "task_abc123",
    "status": "COMPLETED",
    "result": {
      "url": "https://cdn.picsart.com/…/result",
      "items": [
        {
          "url": "https://cdn.picsart.com/…/result"
        }
      ],
      "model": "sam-3-video"
    }
  }
}