CVAT
Image & Video Data Annotation Platform for Computer Vision applications
Quick facts
- Company
- CVAT
- Service type
- Data annotation / AI training data
- Specialties
- image
- Hiring status
- Check the company application page for availability
- Website
- Website
- Careers
- Careers page
- Application guide source checked
- Not yet verified
Application process overview
CVAT (Computer Vision Annotation Tool) is an open-source image and video annotation platform originally built inside Intel in 2016, publicly released in 2018, and now maintained by CVAT.ai Corporation in partnership with OpenCV. It is primarily a B2B tool that teams self-host or use via cvat.ai's hosted version to label data for computer vision, though CVAT.ai Corporation also hires its own in-house data annotation specialists for client projects.
Key findings
Application Process: For CVAT as a tool, users simply sign up at app.cvat.ai for the hosted version or deploy the open-source edition from GitHub. For paid annotator roles, candidates apply via the CVAT Careers page (cvat.ai/careers) for the Data Annotation Specialist position.<br><br>Assessments: Role-specific screening for annotator hires (accuracy tests on images/video/point clouds); no public test for tool users.<br><br>Job Types / Expertise: Bounding boxes, polygons, keypoints/skeletons, cuboids, segmentation masks, 3D point-cloud annotation, and video tracking for computer-vision ML.<br><br>Compensation: Third-party listings (ZipRecruiter) cite a wide $15-$109/hr range for CVAT-skilled annotators hired by outside firms; CVAT.ai's own salaries are not publicly disclosed.<br><br>Flexibility / Remote work: Fully remote roles advertised. The tool itself is browser-based and self-hostable.<br><br>Challenges / Concerns: Most openings are engineering roles, not mass annotator hiring — independent annotators looking for piece-work will find few opportunities directly at CVAT.ai. Hosted version has usage tiers and learning curve for complex tasks.<br><br>Legitimacy: Legitimate; CVAT.ai Corporation (private) is the commercial steward, MIT-licensed open source, backed by OpenCV since 2022.
Conclusion
CVAT is best understood as a tooling company rather than a gig platform: the product is widely trusted by ML teams and is the de facto open-source standard for computer-vision labeling, but opportunities for independent annotators to earn money directly through CVAT.ai are limited to a small number of in-house specialist roles. It suits ML engineers who need a powerful, free annotation tool, enterprise teams who want SSO/RBAC via CVAT Enterprise, and a small pool of experienced annotators seeking full-time employment. Freelancers looking for volume task work should look elsewhere, though CVAT proficiency is a sought-after skill on adjacent platforms.