Labelbox home page
Search...
⌘K
Ask AI
Log in
Start for free
Contact us
Contact us
Search...
Navigation
Foundry
Select data rows
Guides
Developer Guides
Release Notes
Python Tutorials
API reference
Status Page
Community
Blog
Getting Started
Overview
Key Definitions
Technical Requirements & Recommendations
Limits
Customer Support
Product Release Phases
Labeling Services
Overview
Managed Services
Self-service
Alignerr Connect
Alignerr
Annotate
Overview
Labeling Editors
Projects
Code & Grammar Assistance
Import Labels
Best Practices
Quality Analysis
Model
Overview
Fine-tune Model
Experiments
Foundry
Overview
Quick Starts
Foundry Apps
Custom Model Integration
Select Data Rows
Select Prediction Model
Define & Preview Model Run
Track Model Run Progress
View Predictions & Details
Send Predictions to Annotate
Export Predictions
Improve Model Performance
Catalog
Overview
Import & Upload Data
Bulk Classification
Embeddings
Batches
Slices
Filters
Schema
Ontologies
Features
Metadata
Export
Export Labels From Project
Export Data From Catalog
Export Model Run Data
Integrations
Cloud Storage Integrations
Amazon S3
Data Access Audit Logs
Microsoft Azure Blob Storage
Google Cloud Storage
Webhooks
Integration Resources
Manage Team
Overview
Workspaces
Roles & Permissions
Manage Members & Permissions
Monitor Labeling Performance
Workforce Provider RBAC
Notifications
New Member & Group Management Experience
Access & Usage
Manage Account & Billing
Access, Storage, & Security
Single Sign-On (SSO)
Multi-factor Authentication
Updates
Deprecations
Migration Guide: Export V1 to Export V2
Migration Guide: Reporting Page to Monitor
Foundry
Select data rows
Foundry uses foundational models to create prediction(s) or enrich data rows.
To begin using Foundry, use
Catalog
to select the data rows.
As you select data rows, the
Manage selection
menu shows the number of selected rows.
Select
Predict with Foundry
to begin defining your model run.
Next, you will
select the foundational model
to use in your model run.
Custom Model Integration
Select Prediction Model
Assistant
Responses are generated using AI and may contain mistakes.