CVAT

CVAT turns images, video, 3D data, and audio into labeled datasets for AI workflows. Its annotation tools include bounding boxes, masks, polygons, keypoints, skeletons, and object tracks. Teams can review jobs through validation and acceptance stages, and assess quality with ground truth, honeypots, and consensus workflows. The platform connects to Amazon S3, Azure Blob Storage, Google Cloud Storage, S3-compatible buckets, and self-hosted storage. It names SAM 2, SAM 3, Ultralytics, Hugging Face models, and custom models as labeling aids. CVAT exports to more than 20 formats, including COCO, YOLO, KITTI, Cityscapes, and Pascal VOC, with API, SDK, and CLI options for automation. It is available as a managed cloud service or for self-hosted deployment, including on-premises and air-gapped setups. Security features include encryption in transit and at rest, role-based access, SSO, access logging, and audit trails. CVAT says it does not currently hold ISO 27001 or SOC 2 certifications. Free options are available; paid online plans start at $23/user/mo billed annually.

CVAT covers varied annotation types, quality review, and dataset export, with cloud and self-hosted deployment options. Teams needing some enterprise controls should note both Community limits and the stated certification gap.

CVAT suits individuals and teams preparing labeled datasets for AI work, particularly those needing review workflows, storage integrations, or deployment choices. Its Community plan is described for personal use and small teams.

Overview

CVAT is a platform for labeling images, video, 3D data, and audio for AI datasets. It is best suited to teams with varied annotation work and a need for review workflows or deployment flexibility. Its breadth is a clear advantage, but Community leaves out several enterprise-minded controls.

Key features

CVAT supports boxes, polygons, polylines, points, masks, ellipses, cuboids, skeletons, tags, and object tracks. That range is useful for projects that span detection, segmentation, and tracking instead of forcing every dataset into one annotation style.

Teams can move jobs through validation and acceptance stages and assess quality with ground truth, honeypots, and consensus workflows. These options make review part of collaborative labeling, though automated quality assurance is not included in Community.

Storage connections cover Amazon S3, Azure Blob Storage, Google Cloud Storage, S3-compatible buckets, and self-hosted storage. CVAT names SAM 2, SAM 3, Ultralytics, Hugging Face models, and custom models for model-assisted labeling; however, Community lacks SAM integration and built-in AI agent integrations. Exports span more than 20 formats, including COCO, YOLO, KITTI, Cityscapes, and Pascal VOC, while the API, SDK, and CLI support automation around labeling workflows.

CVAT can run as a managed cloud service or be self-hosted, including on-premises and air-gapped deployments. Security controls include encryption in transit and at rest, role-based access, SSO, access logging, and audit trails. It does not currently hold ISO 27001 or SOC 2 certifications, a significant constraint for buyers who require either certification.

Pricing

Community costs nothing and uses the MIT license; it is positioned for personal use and small teams. It gives up automated quality assurance, SSO, SAM integration, built-in AI agent integrations, and advanced reporting, making it less suitable for teams that depend on those capabilities.

CVAT Online Free costs nothing and is capped at 1 member, 1 project, 3 tasks, and 1 GB of internal storage. It is a bounded way for an individual to start, not a roomy team workspace.

CVAT Online Solo Yearly is $23 a year, billed $23 / per month, with a stated 30% saving and an unlimited personal workspace. Solo Monthly is $33 a month for an unlimited personal workspace. These plans suit individual users who want more than the free limits; monthly billing costs more than the annual rate.

CVAT Online Team Yearly is $23 a year, billed $23/per user and $46 / per month for 2 seats, with a stated 30% saving and a 2–50 seat range. Team Monthly is $33 a month, billed $33/per user and $66 / per month for 2 seats, also for 2–50 seats, on a pay-as-you-go basis. The annual option reduces the stated per-user rate, while monthly billing avoids an annual commitment.

Enterprise Basic costs $12,000 a year for a single-instance deployment and essential support for critical issues. It includes 24-hour email support with a 24-hour response SLA and a private helpdesk channel for urgent critical issues. Enterprise Premium has custom pricing and adds advanced deployment, customization, priority expert support, up to 3 instances, and 100 dev hours. Those plans target organizations needing deployment and support beyond the online tiers; Premium's custom price means the cost must be assessed against that scope.

Platforms

CVAT is available through web, API, and self-hosted deployments. The choice of managed cloud, on-premises, or air-gapped installation gives teams deployment options, but organizations with mandatory ISO 27001 or SOC 2 certification requirements should rule it out.

Who it's for

CVAT is a strong fit for teams labeling mixed data types, collaborating on annotation, and exporting to established dataset formats. Its storage integrations and self-hosted options also suit workflows tied to existing infrastructure. Solo users can start with the free tier or choose an unlimited personal workspace; larger teams should compare the Community feature gaps and enterprise support tiers with their requirements.

Pros and cons

• Pros: Broad annotation coverage accommodates detection, segmentation, and tracking in one platform.
• Pros: Validation, acceptance, ground truth, honeypots, and consensus give teams multiple ways to review work.
• Pros: More than 20 export formats, storage integrations, and API/SDK/CLI support make it adaptable to established data workflows.
• Pros: Managed cloud and self-hosted deployment, including air-gapped options, serve different infrastructure needs.
• Cons: Community omits automated quality assurance, SSO, SAM integration, built-in AI agent integrations, and advanced reporting.
• Cons: No ISO 27001 or SOC 2 certification makes CVAT a poor fit when either certification is mandatory.
• Cons: Free online use is tightly capped, while Enterprise Premium requires custom pricing.

Alternatives

Roboflow is another freemium option with a free tier that includes 10 credits a month, enough to train about 30 models or run 80,000 inferences; consider it when that allowance aligns better with your workload.

Labellerr offers a free Researcher Plan with 2,500 data credits, 1 seat, 1 workspace, and up to 10 projects, making it worth considering for students and researchers who fit those limits.

Annota AI is another freemium web and self-hosted option, with a free plan for 1 member, 2 datasets, and 4 GB storage; it may suit an individual whose needs fit that allowance.

Label Studio is a freemium alternative with a free trial and support for API, desktop, web, and self-hosted platforms.

Supervisely has a free Community plan with 5 GB storage, 10,000 files, 2 members, and unlimited projects and annotations; consider it if those caps suit your team.

Ultralytics Platform offers a free plan with 100 GB storage, 10 GB per upload, 100 models, 3 concurrent cloud training jobs, and 3 cloud deployments; it is an alternative for users whose needs align with those resources.

MakeSense.ai is a free, browser-based photo labeling tool.

Vailabel Studio is a free, open-source desktop option that requires no account.

For more options, browse Image Annotation Software, AI Image Annotation Tools, Data Labeling Software, AI Data Labeling Tools, and Video Annotation Software.

Verdict

Choose CVAT if your team needs flexible annotation, structured quality review, and a choice of managed or self-hosted deployment. Its strongest case is breadth across data types and workflow integration; look elsewhere if Community’s missing controls are essential or ISO 27001 or SOC 2 certification is a requirement.