Count Eggs in a Photo — AI Egg Counter

Photograph one visible layer from above in even light, then choose Count. Review every result marker and any dashed reference. Multiple trays can share a photo when each egg remains clear; otherwise count and tally them separately.

Upload the photo and choose Count. Auto mode uses the selected item name; review every marker and correct the tally. For a guided recount, open Examples and mark one complete, clear eggs, with up to three examples for different-looking kinds.

White, brown and speckled eggs can all be estimated, but tray color, glare, shadows and overlap still affect the result. Use even light, then review and correct every marker.

Yes. You can count, review and download a free result without an account. Downloads are free without a watermark. No signup required. Server-processed photos are encrypted in transit and removed under our Security policy.

Counting mode chooses the detection engine behind the count. Similar items is tuned for a tray of eggs that are all one size and color — the everyday case for a collection table — while Distinct objects switches to object detection for separated items in a mixed photo. Auto is a convenient starting point, but review and correct its markers before using the tally.

Sensitivity (Low to High) controls how confident the AI must be before marking an egg. Raise it toward High if pale or brown eggs are blending into a similarly colored tray and getting missed; lower it toward Low if shadows between eggs or the tray's dimples are being counted as extra eggs. Photographing trays from directly above with even light needs the least adjustment.

Result overlay decides how the count is displayed on the photo — Dots + heatmap marks counted detections and shows density, Dots only keeps the annotated photo simple for review, and Heatmap only shows where detections are concentrated. Ignore small detections filters out anything smaller than the threshold you set, useful for reducing cracked-shell or debris markers. Review and correct the result before using it for dispatch.

Yes. Open Bulk audit and every tray or collection-table photo is counted with the Distinct objects detector. Review every result and correct the tally as needed. For an example-guided Auto or Similar items pass, count in Single photo mode one image at a time with your own marked examples.

After counting each tray, tap Add to tally and label the row, for example "Shed 3 — morning collection". Repeat across the day, then export the full tally as a CSV for your production log, generate a PDF stock-take report, or sign in to save it to your account and share a read-only team link for dispatch staff.

Use Distinct objects if the pallet mixes egg sizes or colors so each type is recognised on its own, or Similar items if it's one uniform batch for a faster, more precise count. Keep Sensitivity moderate and raise Ignore small detections slightly to skip any shell fragments, then log each tray's count as its own tally row so the export shows a per-tray breakdown alongside the grand total.

No — switching from Single photo to Bulk audit so you can queue up multiple tray or collection-table photos is a Premium feature; on the free plan you count one photo at a time and add each result to the tally sheet yourself. Single-photo counting, Counting mode, Sensitivity and the Examples setting all stay free.

Yes — the tally sheet's CSV export is free for daily production logs, but generating the branded PDF stock-take report and using "Copy team link" to share a saved session with dispatch staff both require a Premium plan, showing an upgrade prompt on the free plan. Saving a stock-take to your account stays free, though free accounts can only keep a limited number of saved sessions.

Examples are optional in Auto mode and required in Similar items. Upload the tray photo, open Examples and mark 1 to 3 complete, clearly visible eggs, then choose Count when the settings are ready. You can remove and redraw a poor box; review every result marker, especially when the tray color is similar.

Yes, for separate and fully visible items. Choose Distinct objects, then add up to 60 photos on Premium or 200 on Pro; review and correct every photo's detector markers before using one tally per photo plus a grand total. Guided Auto or Similar items cannot reuse one crop across unrelated photos, so process those one at a time with a new example.

No. Bulk object counting requires Distinct objects and never transfers a marked crop to another image. The detector settings apply to up to 60 photos on Premium or 200 on Pro, but every result still needs marker-by-marker review. Temporary uploaded and generated files are processed securely and deleted automatically.

Almost always one of three things. The file is over the 20 MB free upload limit — Premium takes 50 MB and Pro 120 MB. The format is not one this tool accepts, so check the list shown on the upload panel. Or the file is not really the format its name says: renaming a file does not change what is inside it, and that is the reason a valid-looking image gets refused.

It reports what is genuinely in the file, so a blank or surprising result usually means the data was never there — stripped by the app that last saved it, removed by a platform on upload, or absent from the original. Try the untouched original file straight from the camera or source rather than a copy that has been through a messaging app.

Egg Counter supports free standard use within the published processing allowances. Free downloads are clean — no watermark. No signup is required for standard tool use. Photo effects and AI edits have a web-size limit on Free, with the original-resolution result available through Premium or one credit. Crop, resize, rotate, compress and convert stay full size. Premium supports batches of up to 60 files, with the applicable plan limits shown before processing.

No — Egg Counter runs in the browser on desktop and phone. There is nothing to install, no trial to start, and no license to buy before you can see whether it does what you need. Try it on the file you are actually stuck with rather than a sample.