About
What is cvat?
CVAT (Computer Vision Annotation Tool) is an industry-leading, open-source data engine designed for machine learning, specifically for annotating images and videos. It is widely used by teams and companies worldwide to prepare data for AI models, supporting data of any scale. Users can start with a free online version at cvat.ai, which offers up to 10 tasks and 500MB of data, or opt for a self-hosted solution with enterprise support for premium features like SSO, LDAP, and advanced analytics. CVAT supports a wide array of annotation formats and integrates deep learning serverless functions for automatic labeling, significantly speeding up the annotation process. It also provides a Python SDK and CLI for developers.
Best used for
Ideal for data scientists and developers who need to efficiently annotate large volumes of image and video data, prepare high-quality datasets for machine learning, and integrate annotation workflows into their development pipelines. Especially valuable for teams requiring both online accessibility and robust self-hosted enterprise solutions with advanced features.
Common actions
face swapping"AI Agents"github copilotautomated workflowlow-code/no-codecollaborationworkflowsdeepfakeopen-source
Capabilities
Key features
- Interactive video annotation
- Interactive image annotation
- Multiple annotation formats
- Deep learning serverless functions
- Automatic labeling
- Python SDK
- Command line tool
Target Audience
data scientistdeveloperproduct manager
Integrations
roboflowhuggingfacefiftyone
Pricing & Plans
Freemium · Enterprise · Open Source
FAQs
What are the limitations of the free online version of CVAT?
The free online version at cvat.ai allows for up to 10 tasks and 500MB of data. It is suitable for individual use, but it does not include analytics features for team management or the ability to export images, only annotations.
Does CVAT support automatic labeling?
Yes, CVAT supports automatic labeling through deep learning serverless functions, which can significantly speed up the annotation process. It integrates various algorithms and frameworks like PyTorch, OpenVINO, ONNX, and TensorFlow for tasks such as detection, segmentation, and tracking.
Can CVAT be self-hosted?
Yes, CVAT can be set up as a self-hosted solution. Prebuilt Docker images are available for local deployment, and enterprise support offers premium features like SSO, LDAP, Roboflow and HuggingFace integrations, and advanced analytics for self-hosted installations.
Shypd.aiThe largest open index of AI tools
