Loading…
Check a suspicious portrait or video for signs of AI generation and face manipulation. Review the separate findings, keep the source in context and decide what needs another check.
See prices before checkingUse the image detector for a portrait, profile photo or suspicious still. Review AI-generation and available face-manipulation findings separately. An image check does not establish who a person is or who owns the photo.
Check a photoUse the video detector for a clip. Read the available visual, face and synthetic-speech findings separately, together with analysis coverage. A screenshot only preserves one frame and cannot replace a video check.
Check a videoSave the original file when possible, plus the post URL, publication time and claimed source. A repeatedly downloaded or compressed copy can lose details that matter to the analysis.
For video, note the relevant timestamp and check the report's coverage. If you submit a shorter clip, keep its connection to the full recording and avoid changing the material unnecessarily.
Compare the detector findings with the original source and other accounts of the event. If someone asks for money or account access, contact the person through a channel you already trust.
AI generation, face manipulation and synthetic speech are different findings. Read their labels and uncertainty before interpreting a score. An inconclusive result means the available evidence did not settle the question; a low signal does not prove that the media is genuine.
It examines media for patterns associated with synthetic generation or manipulation. A deepfake commonly imitates or changes a real person's face or voice. The detector reports available signals; it does not establish the person's identity or intentions.
Yes. Choose the photo check for a still image or the video check for a clip. Each has its own supported formats, limits and price, shown before you start.
No. A generated landscape or invented character need not imitate a real person. A face swap is a different question from whether the whole image was generated. Review the relevant findings separately.
No. Compression, edits, poor quality and unfamiliar generation methods can affect the result. Use the report with source checks and independent evidence, especially when a decision could affect someone.
No. It can flag available AI or manipulation signals in the image. It does not identify the person, confirm their presence or determine whether the profile has permission to use the photo.
Yes. The image and video API provides recorded findings for your application, while batch checking handles multiple supported images. Your team decides how to review the results and what action to take.