Face Analyzer

Count the faces in a photo and box each one. Works with group shots, selfies, and crowd images. No age, gender or mood profiling — and no identity matching.

3,500+ photos analyzed

Only upload a photo you have the right to use. If someone else is in it, you need their consent. This tool never identifies anyone and never estimates age, gender or emotion — Terms.

Drag & drop a file here, or

JPG PNG WebP GIF BMP AVIF Processed locally in your browser

AI Detection

Neural network trained on millions of faces for accurate detection, even in group shots and crowds.

Group Shots

Counts every face in crowds and group photos, and numbers each one on the annotated image.

No Profiling

Counts faces and nothing else. No age, gender or mood estimation, no identity matching, no data about the people in the photo.

100% Private

Runs entirely in your browser. Photos never leave your device. Only upload photos you have the right to use.

Detection method

The counter uses face-api.js with TinyFaceDetector running entirely in your browser. Upload any image to count face positions — the model identifies them, tells you how many faces are present, and marks each with a bounding box overlay. It processes group photos, crowd shots, and multi-face compositions without uploading images to any server. Detection works across different skin tones, lighting conditions, and face sizes, though very small faces in high-resolution crowd images may need the image to be cropped for best results.

Tips for best results

Front-facing, well-lit photos give the highest accuracy. Faces turned away, heavily occluded, or at extreme angles may be missed by the detector. Very small faces in large crowd photos need higher resolution to be recognized — try cropping the image to focus on specific groups. Good lighting and minimal motion blur improve detection rates. For assessing overall image quality before face counting, pair with the Quality Analyzer.

Common use cases

Event photographers use it to count people in photo group shots for invoicing or coverage verification. Security and compliance teams audit images for the number of individuals present. Social media managers select group photos that include everyone from an event. Teachers verify class photo headcounts. Researchers analyze crowd density in documentation or archival photos. Real estate photographers confirm no unintended faces appear in property listing images.

Privacy considerations

Photos with faces carry inherent privacy risks — they can identify individuals, reveal social connections, and expose locations when combined with metadata. Before sharing group photos publicly, check the Privacy Score to assess what metadata is embedded. Strip identifying data with the EXIF Remover. Our parent photo privacy guide covers specific concerns around sharing photos of children online.

Face counting is a practical tool that goes beyond simple curiosity. In professional photography, knowing the exact headcount in a group shot matters for billing, coverage guarantees, and print sizing. In content moderation, automated face counting helps flag images that may require privacy review before publication. The TinyFaceDetector model is lightweight enough to run in real-time in the browser, making it accessible from any device without installing software. Unlike cloud-based face detection APIs that transmit images to remote servers, Scanly's approach keeps all processing local — an important distinction when handling photos of people. For a broader look at how photos reveal personal information, visit the photo privacy hub. To count faces in images you've already scanned for other data, use the online face counting guide for format and resolution recommendations. No signup or installation is required — just upload and get results. Read our how face detection works for tips, examples, and best practices.

Frequently Asked Questions

How accurate is the face detection?

Our face counter uses TinyFaceDetector, optimized for speed and accuracy. It typically achieves 90%+ accuracy on clear, well-lit photos. Accuracy may decrease with low-resolution images, extreme angles, partial occlusion, or unusual lighting conditions.

Are my photos uploaded to your servers?

No. All face detection happens entirely in your browser using face-api.js. Your photos never leave your device, ensuring complete privacy. The AI model runs locally in your browser.

Can it detect faces in crowd photos?

Yes! The tool works well with group photos and crowd shots. However, very small faces (under 20 pixels) or heavily obscured faces may not be detected. For best results with large crowds, use high-resolution images where faces are clearly visible.

Why is the first scan slow?

The first scan downloads the face detection AI model (about 200KB). This is cached in your browser, making subsequent scans much faster. The model only needs to download once per browser session.

Does this tool identify or recognize specific people?

No. This tool only counts and locates faces—it does not perform facial recognition or identify individuals. It cannot match faces to names or compare faces across different photos. It's purely a counting and detection tool.

Can I download the annotated image?

Yes! After detection, click "Download Annotated" to save a copy of your image with green boxes drawn around each detected face. Each face is numbered for easy reference. Great for documentation or verification purposes.

What image formats are supported?

The face counter supports JPG, PNG, and WebP images. There's no file size limit, but very large images may take longer to process. For fastest results, images under 4000x4000 pixels work best.