Back to Ultralytics

Roboflow Dataset Import - Ultralytics Platform

docs/en/platform/integrations/roboflow.md

8.4.1213.4 KB
Original Source

Roboflow Integration

The Roboflow integration imports every supported dataset in your Roboflow workspace into Ultralytics Platform at its latest version. Re-run it any time to pull in datasets you've added since your last import.

Import from Roboflow

  1. Go to Settings > Integrations and select Roboflow from the integration list.
  2. Paste your Roboflow API Key and click Import.
  3. Review the Import from Roboflow preview dialog, which lists:
    • New datasets that will be imported
    • Already imported datasets that will be skipped
    • Any datasets with a missing version, unsupported tasks, or that couldn't be sized
    • Storage required, checked against your remaining storage
  4. Click Import to start.

<!-- screenshot --> Imported datasets appear in your Datasets list immediately with a processing status and become ready once their images and annotations finish importing.

Any workspace editor can run the import. The preview is the slow step: Roboflow generates each new dataset's export archive before it can report a size, so a workspace with many fresh versions can take a minute or more to preview. If a very large workspace times out, run it again — everything already imported is skipped, so the second pass only has the remainder to do.

Supported Task Types

Roboflow projects are mapped to the matching YOLO task type:

Roboflow Project TypePlatform Task
Object DetectionDetect
Instance SegmentationSegment
Single-Label ClassificationClassify
Keypoint DetectionPose

!!! note "Where to find your Roboflow API key"

Follow Roboflow's guide to [find your API key](https://docs.roboflow.com/developer/authentication/find-your-roboflow-api-key). Platform uses the key to run the preview and the import and then discards it — nothing is saved, and there is no connection to disconnect later.

!!! note "Unsupported projects are skipped"

Projects with unsupported task types (such as multi-label classification and semantic segmentation) and projects that don't yet have a generated version are skipped and reported in the preview. A dataset whose export size Roboflow can't report in time is also left out and shown as **Couldn't size — retry later**.

!!! tip "Re-run to sync new datasets"

Already-imported dataset versions are detected and skipped automatically, so you can safely re-run the import to pull in newly added datasets without creating duplicates. Bumping a project's version in Roboflow makes it a new dataset to import rather than an update to the one you already have.

!!! warning "Trashed imports still count as imported"

A previously imported dataset that you moved to [Trash](../account/trash.md) is still detected and skipped, because it keeps occupying storage during the retention window. Delete it permanently before re-importing the same Roboflow version.