VGG Image Annotator

VGG Image Annotator (VIA) is a free suite for manually labeling images, audio and video. It best suits researchers and teams that need annotation tools that can run in a browser without installation.

Overview

VIA is distributed as a single HTML file and runs offline in a modern browser. Built with HTML, CSS and JavaScript, it has no external libraries or setup requirements. That makes it a practical choice for standalone annotation work, but it is not a model-assisted labeling tool.

Choose between two versions: VIA3 supports images, video and audio, including shared projects; VIA2 handles images but does not support shared projects and remains maintained. For collaboration, VIA tools can use VGG's public free Version File Store (VFS), or users can run their own.

Key features

Annotators can mark regions with rectangles, circles, ellipses, points, polygons and polylines. Projects can import and export CSV or JSON annotations. Conversion scripts support COCO and Web Annotation formats, and a Python package converts VIA projects to COCO. These export routes help fit VIA into annotation workflows, though the tool itself is oriented toward manual labeling.

The makers report testing VIA in Firefox, Chrome and Safari, and say users have found it works in other browsers. ZIP archives and standalone HTML copies are available for versions 2 and 3. The in-app guide, FAQ and email contacts offer ways to get help. VIA is released under the permissive two-clause BSD license for academic and commercial use.

Pricing

VGG Image Annotator costs nothing. The free plan has no paid tiers or usage limits stated, and VIA has no free trial because it is free to use. There are no stated seat or quota caps. Its permissive license also allows commercial use, so it suits both individual researchers and organizations that need manual annotation without a subscription.

Platforms

VIA is available for Linux, macOS, Windows and the web. Its single-file browser format lets users run it offline without installation or setup. The project describes browser testing in Firefox, Chrome and Safari, but does not claim security certifications or compliance. Offline operation may suit work that should not require a network connection, but it should not be mistaken for a formal security assurance.

Who it's for

VIA is a strong fit for academic projects and commercial applications that need manual image, audio or video annotation, especially when offline use and straightforward exports matter. VIA3 is the better fit for teams sharing projects across media types. VIA2 remains an option for image-only work. Look elsewhere if model-assisted labeling is essential.

Pros and cons

• Pros: Free use with no stated usage limits and a permissive BSD license makes VIA accessible for academic and commercial projects.
• Pros: Offline, single-file browser apps avoid installation and support disconnected annotation work.
• Pros: CSV, JSON, COCO and Web Annotation routes give projects multiple ways to move into other workflows.
• Cons: VIA does not offer model-assisted labeling, so teams seeking AI-assisted annotation will need another tool.
• Cons: VIA2 does not support shared projects; collaboration requires VIA3 and a VFS.
• Cons: No security certifications or compliance claims are made, which may matter for organizations with formal requirements.

Alternatives

Browse Image Annotation Software for more options, or compare AI Image Annotation Tools, AI Video Search Tools, AI Data Labeling Tools and Data Labeling Software if your needs extend beyond manual VIA annotation.

Label Studio is worth considering if you want a freemium option with a free trial and web, desktop, API and self-hosted platforms. MakeSense.ai is another free browser-based photo labeling tool. Roboflow may suit users looking for a freemium tool with a free tier of 10 monthly credits, described as enough for roughly 30 model trainings or 80,000 inferences. CVAT offers a free Community plan for personal use and small teams, as well as an online free plan capped at one member, one project, three tasks and 1 GB. Vailabel Studio is a free, open-source option for Linux, macOS and Windows that requires no account. Annota AI may fit a single member who can work within its free plan's two datasets, 4 GB storage and €4 monthly credits. Labellerr offers researchers a free plan with one seat, one workspace, up to 10 projects and 2,500 data credits. Segments.ai has a free academic license for non-commercial research, with a license request required.

Verdict

Choose VIA if you need free, offline-capable manual annotation with flexible exports and a permissive license. Its biggest advantage is that the browser-based tools work without installation; its main limitation is the absence of model-assisted labeling. Teams needing shared work should use VIA3 with a VFS, while those seeking AI-assisted workflows should look elsewhere.