CVAT vs Encord
A side-by-side comparison of CVAT and Encord, two Vision tools, drawn from Ignaite's continuously-verified listings.
Compared from listings verified as of
At a glance
| Attribute | CVAT | Encord |
|---|---|---|
| Category | Vision | Vision |
| Pricing (differs) | FREEMIUM | PAID |
| License (differs) | Open core | Proprietary |
| Deployment | Hybrid | Hybrid |
| Platforms | Web, API | Web, API |
| Model support (differs) | Model-agnostic | Multi-model |
| Vendor (differs) | CVAT.ai | Encord |
| Capabilities (differs) |
|
|
The honest brief
CVAT
One of the most widely deployed open-source CV labeling tools — millions of self-hosted installs, deep video tooling, and free MIT self-hosting.
- Boxes, polygons, keypoints, 3D, and video
- SAM 2/3-assisted auto-labeling
- Self-host, hosted Online, or Enterprise tiers
- Large community: 15k+ GitHub stars
- Self-hosting needs Docker ops effort
- Annotation only — no model training built in
Encord
Labels DICOM, NIfTI, LiDAR and SAR alongside images/video — built for regulated medical and physical-world AI.
- DICOM/NIfTI/point-cloud support
- HIPAA/SOC 2 for regulated data
- Annotate + curate + index in one
- Model-assisted labeling (SAM, GPT-4o)
- Enterprise pricing, no free tier
- Heavier than lightweight labelers
- Onboarding/setup overhead
- Overkill for simple image tasks
When to pick which
Across the signals we compare, CVAT and Encord differ on pricing and license:
- Pricing
- FREEMIUM · PAID
- License
- Open core · Proprietary
Their capability lists differ in recorded depth — compare the full lists above before deciding.