Beyond Vision: CVAT Goes Multimodal and Introduces Audio Annotation. Learn more
The open-source annotation platform thousands of teams already run, now built for enterprise scale and security.
Trusted by ML teams worldwide
Read customer storiesEvery modality, every annotation type, from first label to production pipeline.
Label images, video, 3D point clouds, and audio in one place, with quality controls built in.
Explore annotation & labelingQuality control for Task #42
Completed requirements
3 of 5
0.0%
Cut labeling time by up to 80%. Pre-label with SAM, YOLO, Hugging Face, or your own models, then let your team verify.
Explore AI automation
AI Tools
Label
Interactor
Convert masks to polygons
Start with a bounding box
Automate task creation, sync cloud storage, and stream verified data into training, all via Python SDK and REST API.
Explore Python SDK & APIAdd vetted annotators to your own CVAT instance for heavy labeling sprints, with no hiring.
Explore Annotation Services
Deploy where your data lives, with the controls your security team requires.
Run CVAT on-prem, in your private cloud, or fully air-gapped.
SSO and role-based access control across teams and projects.
Access logging and audit trails for every action.
GDPR, CCPA, and EU AI Act compliant. Encrypted in transit and at rest.
Sharper motion tracking and scene understanding for safer autonomous trucks.
Read the storyStart with the open-source core, or talk to us about running CVAT at enterprise scale.
