Choose the files that belong together
Use a batch for one claim, a set of product listings or a group of profile photos. Keep the original copies in your own system and use filenames that let you match each report to its source.
Loading…
Check incoming photos and videos together with our own DeepfakePolicy system. Select files or a ZIP archive, follow each analysis and download the reports for your case, client or team.
Create a company workspace, choose your files and review the cost. Add a payment card before starting the checks.
A result and a report for each file, with a summary to download.
| File | AI generation signal | Status | Report |
|---|---|---|---|
| demo-portrait.jpg | 82%Strong AI indicators | Completed | |
| demo-profile-02.jpg | 12%Few AI indicators | Completed | |
| demo-profile-03.jpg | 41%Review needed | Completed |
Example PDF reports are in English. Your own reports are available after processing.
Use a batch for one claim, a set of product listings or a group of profile photos. Keep the original copies in your own system and use filenames that let you match each report to its source.
Upload JPG, PNG, WebP or GIF images, or select them from a ZIP archive. Check the file list and remove anything unrelated. Unsupported files and duplicates are identified during preparation; a ZIP does not turn them into supported images.
The list keeps the filename, detector finding and report together. Investigate unexpected or conflicting signals, then download the individual PDFs, a CSV summary or a ZIP containing the completed reports.
Each image is checked separately. One suspicious photo does not turn every other file in the batch into a positive result.
Read the recorded finding, available file details, image preview when available, interpretation and limitations. The PDF preserves the analysis record beyond the dashboard.
Use the CSV to reconcile filenames, statuses and findings with your own records. A failed or unavailable result remains distinct from a completed low detector score.
Use the browser workspace when a person selects the files and reviews the results. A company API key is not required for this workflow.
Check a batchUse the API when your own application receives the uploads. Submit one job per file, collect structured reports and handle completion through polling or signed webhooks.
Explore the detection APIJPG, PNG, WebP: up to 50 MB · GIF: up to 20 MB, first frame only · MP4/MOV: up to 20 MB and 60 seconds · up to 50 files, 200 MB total
JPG, PNG and WebP: up to 50 MB per file. GIF: up to 20 MB; only the first frame is analyzed. Company batches can also include supported short MP4/MOV videos. The batch limit applies to the selected, expanded files as well as the archive size.
Create a business workspace and connect company billing before processing real files. API and batch usage share spending controls. Completed checks are charged, including inconclusive reports; processing failures are not.
Reports are available for 7 days. Download your CSV and PDF files before they expire. Originals are removed after processing; report thumbnails expire with the report.
Yes. The importer lists supported images from the archive so you can review the selection. The file-count and expanded-size limits still apply. Unsupported archive entries are skipped or reported.
No. The detector provides signals about the submitted files. A genuine image may be reused out of context, and a local edit may escape detection. Keep the final decision with the person reviewing the full case.
Keep the tab open until the company uploads finish. Processing then continues on the server and you can return to the batch from your account. Files that did not finish uploading must be selected again.
No. The sandbox returns fixed examples for integration testing. A batch processes the files you selected and requires active company billing. No sales call is required.