Skip to content
ImageLabeling

Image annotation tool with AI pre-labels

Upload an image. AI draws the boxes, masks and points, you fix them in seconds, then you export COCO, YOLO or Pascal VOC.

Drop an image here or click to upload JPG, PNG or WebP, up to 8 MB.
Pallet rack in a logistics warehouse with cardboard boxes, a forklift in the aisle and a worker in a high visibility vest and hard hat

Or try one of these images

Warehouse shelf, labeled by AI

Classes

ObjectClassConfidenceBox

Unlock every object

Create your account to see and edit every object, then export this image in COCO, YOLO or Pascal VOC.

Drag a corner to fix a box, click once to rename a class, then export the whole image.

Image annotation built for your industry

Pick your industry and see the label schema, annotation types and export format that fit it.

See all use cases

Every annotation type your dataset needs

Each tool has a keyboard shortcut. Every label exports to the same 4 open formats.

COCOYOLOPascal VOCCSV
B

Bounding boxes

Tight rectangles around each object for detection models. Shortcut B. AI drafts them, you drag the corners.

Bounding box annotation
P

Polygons and masks

Pixel accurate outlines for segmentation, drawn with polygon and brush tools. Shortcut P.

Image segmentation tool
K

Keypoints

Points and skeletons for pose and landmarks, with a visibility flag per point. Shortcut K.

Keypoint annotation tool
AD

Classification

One or more classes per image, with attributes, for classifiers and data filtering. Shortcut A accepts, D moves to the next image.

Data annotation tools

The image annotation workflow, from upload to export

Five steps take a folder of raw photos to a versioned dataset your training script can read.

See how it works

  1. 1. Upload

    Drag and drop images or a zip, or import from a cloud bucket URL.

  2. 2. Pre-label

    The AI model proposes boxes, masks, keypoints and classes for your label schema, with a confidence score on every object.

  3. 3. Fix

    Your team accepts, edits or rejects each object in a keyboard first editor.

  4. 4. Review

    A reviewer approves the image or sends it back with a comment. QA tracks who labeled what.

  5. 5. Export

    Freeze a dataset version, split it into train, validation and test, and export COCO, YOLO, Pascal VOC or CSV.

A data annotation platform you can put in the budget

Labeling is a standing cost line for every vision team. ImageLabeling keeps it predictable and auditable.

1 Flat monthly price per workspace, so the cost per labeled image falls as your team gets faster.
2 AI draws the first draft, so your people spend their time on corrections, not on clicking every box.
3 A review trail on every image: who labeled it, who approved it, what changed.
4 Your images are used only to label your dataset and never to train third party models.
5 Google and Microsoft SSO and an audit log on Business, SAML SSO, a DPA and a 99.9 percent uptime SLA on Enterprise.
6 Invoices on yearly plans and open export formats, so you are never locked in.
Read about the data annotation platform Read about data security
Head of a computer vision team at a standing desk reviewing labeled warehouse images on a large monitor

Image annotation tool pricing

Save 50 percent with yearly billing. Every plan includes COCO, YOLO, Pascal VOC and CSV export.

Save 50 percent

Starter

24 USD per month, billed yearly (288 USD)

49 USD per month, billed monthly

3 seats, 1,000 AI pre-labeled images per month, 10,000 stored images.

Team

74 USD per month, billed yearly (888 USD)

149 USD per month, billed monthly

10 seats, 3,500 AI pre-labeled images per month, segmentation masks, full review workflow, API.

Business

249 USD per month, billed yearly (2,988 USD)

499 USD per month, billed monthly

30 seats, 12,000 AI pre-labeled images per month, keypoint skeletons, QA scores, webhooks, SSO, audit log.

Enterprise

749 USD per month, billed yearly (8,988 USD)

1,499 USD per month, billed monthly

Unlimited seats, 35,000 AI pre-labeled images per month, SAML SSO, DPA, 99.9 percent uptime SLA.

Compare all plan details

How ImageLabeling compares with the usual ways to label

Choose a labeling approach around who drafts the labels, who checks them and how the dataset reaches your training code.

PointImageLabelingManual in-house labelingOutsourced vendorOpen source tool
First draftAI proposes labels for your team to edit.Your team draws the labels.The vendor prepares labels to your brief.Manual or assisted labeling depends on the tool and setup.
Review and QAReviewer approval and a label history.Your team defines and runs the checks.Review steps depend on the agreement.Review features depend on the tool.
ExportCOCO, YOLO, Pascal VOC and CSV on every plan.Formats depend on your team's tools.Delivery formats depend on the agreement.Supported formats depend on the tool.
CostFrom 49 USD monthly or 24 USD per month billed yearly.Staff time and tooling.Service fees depend on scope.Hosting, setup and maintenance need resources.
Your dataImages never train third party models.Your team manages access and storage.Data handling depends on the agreement.Your team manages data when self hosting.

Read the full comparison to choose a workflow for your dataset.

Annotator at a home desk drawing outlines on a photo with a pen tablet, second monitor showing a list

Vetted annotators who work inside your workspace

Request annotators who follow your label schema and pass your QA rules. Available on Business and Enterprise as a monthly retainer from 899 USD per dedicated annotator.

  • 1 Annotators pass a qualification task before they touch client data.
  • 2 They work in your projects with the role you give them, and you see every label they make.
  • 3 Your reviewers stay in control of what reaches the export.

Explore data annotation services

Experienced annotator looking for work? Apply to join the workforce.

Image labeling guides

Practical reading for teams building their first dataset or their fiftieth.

Read the guide

Questions teams ask before they pick an image annotation tool

See all questions

Is my data used to train AI models?

No. Your images are processed only to label your dataset. They are never used to train third party models, and you can delete them at any time.

Which formats can I export?

COCO JSON, YOLO txt files with a data.yaml, Pascal VOC XML and CSV, on every plan.

Can I import labels I already have?

Yes. Import existing labels in COCO, YOLO, Pascal VOC or CSV and keep working on them in the editor.

How accurate are the AI pre-labels?

They are a first draft with a confidence score on every object. Your team accepts or fixes each one, and with review turned on nothing reaches the export without a reviewer's approval.

What counts toward the AI image limit?

One image pre-labeled by AI once counts as one. Manual labeling and re-exports do not count. Deleting images never frees the monthly AI quota.

Can several people label at the same time?

Yes. Every plan includes seats and roles: owner, admin, reviewer, labeler and viewer, with access per project.

Do you offer people to do the labeling?

Yes. Vetted annotators are available on Business and Enterprise as a monthly retainer, working inside your workspace under your QA rules.

Can I try ImageLabeling before I pay?

Yes, the free demo at the top of this page labels one image. Plans start at 24 USD per month billed yearly.

Can I cancel?

Yes, anytime. Access runs to the end of the paid period and payments are final, as described in the refund policy.

Does ImageLabeling label video?

Upload video frames as images and label them like any other image. A native video timeline is not part of the product.

Label your first image now

Upload one image and see boxes, masks or keypoints drawn in seconds. Create your account to label the rest of your dataset.

Drop an image here or click to upload JPG, PNG or WebP, up to 8 MB.
Label my image