Built by Metorial, the integration platform for agentic AI.
Provider Summary
discover projects and image batches
upload and manage images
annotate image datasets
train computer vision models
run object detection inference
create and export dataset versions
manage annotation jobs
search images visually
Manage Roboflow computer vision projects and image datasets. Discover projects and image batches, upload and annotate images, search and tag datasets, generate and export dataset versions, start and stop model training, list trained models, run hosted object detection inference, and manage annotation jobs.
Connect with OAuth or a workspace-scoped private API key. Call who_am_i to verify the connected workspace and list_projects to discover project IDs. Tools accept project slugs or the full workspace/project IDs returned by discovery. Dataset exports produce downloadable ZIP files. Training runs asynchronously and can consume credits; run_inference uses a deployed object-detection model.
Register an app under Workspace Settings > Developer with the exact callback URI, token authentication set to client_secret_post, and Unlisted visibility for connections from other workspaces. Enable the scopes requested by this integration on the app. Customer workspace policy may require allowing the client ID.
OAuth uses PKCE and refresh tokens, and discovers the authorized workspace automatically. It requests workspace/project reads and project creation, image upload/read/tag/annotation, dataset version creation/read, training and annotation jobs, batches, and inference permissions. It retains API-key authentication for existing connections.
Provider documentation confirms OAuth for the REST API. The published scope list has no project/image deletion scopes; OAuth deletion support and hosted inference have not been verified live. Use a private API key when these operations are required.
Provider references: OAuth setup and scopes, developer reference, REST authentication.
This integration is licensed under the FSL-1.1.
Built with ❤️ by Metorial