Skip to main content
Auto-label jobs allow you to automatically generate annotations using machine learning models. Configure confidence thresholds, select specific labels, and run dry runs to preview results before applying them to your project.

​
List auto-label jobs

Retrieve a paginated list of auto-label jobs. 
const jobs = await avala.autoLabelJobs.list({
  project: 'project-uid',
  status: 'completed',
  limit: 50,
  cursor: 'next-cursor'
});

​
Parameters

​
options.project
string
Filter by project UID
​
options.status
string
Filter by job status (e.g., ‘pending’, ‘running’, ‘completed’, ‘failed’)
​
options.limit
number
Maximum number of jobs to return per page
​
options.cursor
string
Cursor for pagination

​
Returns

Returns a cursor-paginated list of auto-label job objects.
​
items
AutoLabelJob[]
Array of auto-label job objects
​
nextCursor
string | null
Cursor for the next page of results
​
previousCursor
string | null
Cursor for the previous page of results
​
hasMore
boolean
Whether more results are available

​
Get an auto-label job

Retrieve a specific auto-label job by UID. 
const job = await avala.autoLabelJobs.get('job-uid');

​
Parameters

​
uid
string
required
The unique identifier of the auto-label job

​
Returns

Returns the auto-label job object.
​
uid
string
Unique identifier for the job
​
status
string | null
Current status of the job
​
modelType
string | null
Type of model used for auto-labeling
​
confidenceThreshold
number | null
Minimum confidence threshold for predictions
​
labels
string[]
Labels to auto-label
​
dryRun
boolean
Whether this is a dry run (preview only)
​
totalItems
number
Total number of items to process
​
processedItems
number
Number of items processed so far
​
successfulItems
number
Number of items successfully labeled
​
failedItems
number
Number of items that failed
​
skippedItems
number
Number of items skipped
​
progressPct
number | null
Progress percentage (0-100)
​
errorMessage
string | null
Error message if the job failed
​
summary
object | null
Summary of results
​
startedAt
string | null
ISO 8601 timestamp of when the job started
​
completedAt
string | null
ISO 8601 timestamp of when the job completed
​
createdAt
string | null
ISO 8601 timestamp of when the job was created

​
Create an auto-label job

Create a new auto-label job for a project. 
const job = await avala.autoLabelJobs.create('project-uid', {
  modelType: 'yolov8',
  confidenceThreshold: 0.7,
  labels: ['person', 'car', 'bicycle'],
  dryRun: true
});

​
Parameters

​
projectUid
string
required
The unique identifier of the project
​
options.modelType
string
Type of model to use for auto-labeling (e.g., ‘yolov8’, ‘sam’)
​
options.confidenceThreshold
number
Minimum confidence threshold for accepting predictions (0.0 to 1.0)
​
options.labels
string[]
Specific labels to auto-label. If not provided, all labels will be used
​
options.dryRun
boolean
If true, preview results without applying annotations

​
Returns

Returns the created auto-label job object.

​
Cancel an auto-label job

Cancel a running auto-label job. 
await avala.autoLabelJobs.cancel('job-uid');

​
Parameters

​
uid
string
required
The unique identifier of the job to cancel

​
Returns

Returns void on success.

Build docs developers (and LLMs) love

Get started for freeTalk to us