AI polygons keep the right amount of detailv7.2.0

From images to training data,on your machine.

AI-assisted annotation with SAM2, SAM3, YOLO-World, and more, all running locally.Your images never leave your computer.

Try Labelme free
See plans

macOS, Windows, and Linux·Free 3-day trial · One-time purchase from $49 · or install open-source Labelme

By the creator of open-source Labelme · 16K+ GitHub stars · 4.9★ from 146 reviews · 920+ paid licenses

The Labelme desktop app annotating a synthetic Urban Plaza dataset: a selected dog polygon and the label dialog naming it, shown enlarged in a loupe, with drawing tools, label lists, and dataset files around the canvas.

Agriculture, medical, industrial, traffic, and more

What people say about Labelme.

Verified Gumroad reviews

5 out of 5 stars

"The best standalone annotation tool."

Royi AvitalVerified purchase
5 out of 5 stars

"I used the free version before, and saved a lot of hassle, now, this wouldn't be much to invest."

Yoon Young JoonVerified purchase
By the creator of open-source Labelme (16K+ stars on GitHub)4.9★ from 146 verified reviews920+ paid licenses

From raw images to a training-ready dataset.

Open a folder, outline objects, check the labels, and export YOLO and Pascal VOC files ready for training.

Labelme · orchard_survey_0040.jpgStage 02 · Outline
Every tree crown outlined with an editable polygon in Labelme, the farm truck marked in amber, and each shape listed in the annotation list.
Editable polygons · local SAM2 + SAM3See dataset tools

Your images stay on your computer.

Annotate with local AI without sending source images, annotations, or exports to a cloud service.

  • DataNothing is uploaded. Source images, annotations, and exports stay on disk.
  • Local AIRuns on your workstation with SAM2, SAM3, and YOLO-World. No API key or third-party AI account.
  • SetupDownload the app and each model once, then work offline.
Read privacy and AI details
Offline
Local AI
  • SAM family
  • YOLO-World
Labelme annotating a synthetic medical image on the desktop.
Stays local
  • Annotations
  • Exports

Point, box, or words. AI drafts the annotation.

Three ways to prompt the local models. Every result stays editable, point by point.

Included in Starter + ProRuns locally
See plans →
One box promptMultiple editable instances
A Labelme parking-lot view with SAM3 oriented boxes around every visible car.
Draw one rough box on one car; SAM3 returns an oriented box for every car. Each result stays editable.

Everything Labelme does.

Both paid tiers start from the open-source app: they add the packaging around it and the tools that turn annotations into training data.

  • In open source

    Every annotation feature: polygons, boxes, masks, and the SAM2, SAM3, and YOLO-World models. The paid tiers start from this same app.

  • What Starter adds

    The tested desktop build: no pip, no dependency management. Model downloads route through the Labelme mirror, so they arrive even behind a corporate firewall.

  • What Pro adds

    The dataset toolkit, ready-to-train exports (YOLO, YOLO-OBB, VOC), and priority email from the author.

Runs on macOS, Windows, and Linux · app in 14 languages

Compare Starter and Pro

Same Labelme. Less setup.

Install the open-source Python package, or download the official desktop app for macOS, Windows, and Linux.

wkentaro / labelme16K+ GitHub stars

The trusted annotation foundation, created and maintained since 2016.

A real macOS desktop showing Labelme annotating an urban scene, with source images, exports, and the Labelme app in the Dock.
Real Labelme workflow · local AI · source images and exports stay on your desktop
Open-source installOfficial desktop build
InstallPyPI or a Linux packagePrebuilt app for macOS, Windows, and Linux
Model downloadsGitHub and Hugging FaceLabelme mirror first, with GitHub and Hugging Face as backup
Updates & supportSelf-managed updates and community supportOne year of app updates. Pro adds priority support.
Try Labelme free

3-day trial · Runs offline · One-time purchase · 14-day money-back guarantee

Three plans. One number each.

One-time purchase, yours to keep. Every plan runs fully on your machine.

Starter

Desktop build

Installs without Python, pulls models through the Labelme mirror even behind a corporate firewall, one year of updates

$49
Get Starter

Pro (Lifetime)

Updates forever

Everything in Pro, plus lifetime upgrades including future major versions

$249
Get Lifetime

3-day trial · Runs offline · One-time purchase · 14-day money-back guarantee

Compare all plans →

Prefer the open-source route? The labelme Python package stays available on PyPI, no purchase needed.

Questions, answered plainly.

Everything worth checking before you bring Labelme into your annotation workflow, from local AI and export formats to licensing and refunds.

Choosing Labelme

What is Labelme?

Labelme is an image annotation tool for creating training datasets. The open-source route is a Python application. Starter and Pro package that foundation as a standalone desktop app for macOS, Windows, and Linux, with AI models that download once and then run locally.

Is Labelme open source?

Yes. The annotation application is GPL-3.0 and maintained on GitHub, and that same app runs inside the paid builds. Starter adds the tested binary and the model mirror. Pro also adds the dataset toolkit, which is closed source.

Is this the same as MIT CSAIL LabelMe?

No. Labelme by Kentaro Wada is distinct from the MIT CSAIL LabelMe project. This site is for the modern open-source application and its paid offline desktop build.

What’s the difference between open source, Starter, and Pro?

The open-source installation gives you the Python application and its annotation workflow. Starter is the tested desktop build, so there is no Python setup, and model weights arrive through our own mirror that works behind restrictive firewalls. Pro adds the dataset toolkit, ready-to-train exports, advanced guides, and priority email support. Compare the plans.

Privacy and AI

Does Labelme work without an internet connection?

Yes, after a one-time download per model. Each AI model is fetched on first use, and after that annotation and AI assistance run locally with no connection. If your network filters outbound traffic, allow osam-models.labelme.io on port 443. The Starter and Pro builds fetch models from there first.

Can I run Labelme fully air-gapped?

Yes, with one step. Download the models on a connected machine first, or ask support@labelme.io for the offline pack.

Do my images ever leave my computer?

No. Images, annotations, and exports stay on your machine, with no upload step, no cloud processing, and no telemetry. The app downloads model weights on first use, and nothing else leaves the workstation.

Which AI models are supported?

Labelme supports the SAM family (SAM, SAM2, SAM3, and EfficientSAM) for assisted segmentation and YOLO-World for text-prompt object detection. The models run locally rather than through a hosted API. In Labelme, SAM3 is driven by box and text prompts, so use SAM2 or EfficientSAM for point prompts. YOLO-World returns bounding boxes.

Do I need an API key or cloud account?

No API key, no third-party AI account, and no activation or sign-in inside the app. You do need a free labelme.io account to download the app and its updates.

Annotation workflow

Which operating systems are supported?

The desktop app runs on macOS (Apple Silicon), Windows (64-bit), and Linux (64-bit). Intel Macs are supported up to v7.0.4. The open-source application installs through Python anywhere Python runs.

What annotation types can I create?

Labelme supports polygons, rectangles, oriented rectangles, circles, lines, points, masks, and rich attributes. AI assistance can propose masks and polygons while keeping the final geometry under your control.

Which export formats are available?

Pro includes automated conversion for YOLO, YOLO-OBB, and Pascal VOC, plus import back from YOLO and YOLO-OBB. The toolkit also produces masks, visualizations, and dataset statistics from the command line.

What does the Pro dataset toolkit add?

The toolkit provides utilities for batch processing, visual review, resizing images together with their annotations, label renaming, dataset statistics, export, and CLI automation.

Plans, purchase, and licensing

Can I try Labelme before buying?

Yes. The desktop app has a three-day trial of the full annotation app, with no credit card, and you can download it here. The trial is tracked locally, so a new labelme.io account does not start it over, and the Pro dataset toolkit is not part of it. The open-source package remains available through Python.

Is Labelme a subscription?

No. Every plan is a one-time purchase with no recurring fee. Starter and Pro include one year of updates, and the version you have keeps working after that. Pro Lifetime includes every future update.

What happens after I purchase?

Access is granted as soon as payment clears. Pay with the same email you want to sign in with. If you do not have an account yet, the purchase creates one and sends a link to set a password. Then sign in to download the app.

What happens when my year of updates ends?

Nothing stops working. There is no license check and no expiry in the app, so the version you have keeps running indefinitely. You only pay again if you want newer releases, or choose Pro Lifetime to get all of them.

What is the refund policy?

Fourteen days, no questions asked. Email support@labelme.io within fourteen days of payment and we refund in full through Paddle to your original payment method. A refund ends your access.

Can I use Labelme commercially?

Yes, for your own or your organization’s internal annotation and dataset work. The annotation app is GPL-3.0. The Pro toolkit and other paid features are licensed to one named user and may not be redistributed, resold, or sublicensed.

What licenses apply to the app and the AI models?

The annotation application is GPL-3.0. The AI models carry their own licenses: SAM, SAM2, and EfficientSAM are Apache-2.0, SAM3 is under Meta’s SAM License, and YOLO-World is GPL-3.0. Contact support@labelme.io if your organization needs the license texts for review.

Can I use my license on more than one computer?

Yes. A license covers one person, not one device, so you can install it on as many of your own machines as you use. There is no activation step and no device limit.

Can I purchase Labelme for a team or organization?

Yes. Team licenses cover five named seats, and larger seat counts, purchase orders, and invoices are arranged by email. Contact support@labelme.io and we will set it up. Academic, institutional, and reseller purchases are supported. Each seat belongs to one named person, so shared accounts are not permitted.

Can I receive an invoice or claim tax exemption?

Paddle sends a receipt and invoice after purchase. Enter your VAT number at checkout to apply EU reverse charge, because it cannot be applied afterwards. Buyers elsewhere can send exemption documents to assist@paddle.com.

Your next dataset stays yours.

Try Labelme on your machine, then choose the plan that fits your work.

Try Labelme free
See plans

Available for macOS, Windows, and Linux·Free 3-day trial · One-time purchase from $49 · 14-day money-back guarantee