How AI Scan works
- Add a video or photo. Videos up to 10 minutes, or a JPG, PNG or WebP photo.
- Scan. AI looks at 9 moments across the video and listens to the speech.
- See the list. People, objects, places, activities, text and what’s said — download it as text or JSON.
Useful for
- Video descriptions · know what’s in your clip
- Tags & keywords · objects and places for SEO
- Accessibility · describe photos and videos
- Checking footage · find logos and visible text
- Notes · what was said, with timestamps
Frequently asked questions
What does AI Scan find in a video?
It lists the people (how many, described by appearance — never identified), objects, animals, places, activities and any text it can see, with a one-line summary. For videos with sound it also gives the speech with timestamps, the spoken language and a sound-level timeline.
Can I scan a photo?
Yes. Add a JPG, PNG or WebP photo and AI Scan lists everything it can see in it.
Does it recognise who people are?
No. People are only described by appearance and what they are doing. AI Scan never names or identifies real people.
How long can the video be?
Up to 10 minutes. AI Scan looks at 9 moments spread across the video; trim longer videos first.
Is my video stored?
No. Only 9 still frames (or your photo) and a small speech-only copy of the sound are sent to our AI provider, and nothing is kept.
Is AI Scan free?
A scan uses 2 credits, returned if it can’t be done. New accounts get starting credits and you can earn more by sharing Chapalta Studio. It’s free during the beta.
Also try AI Voice Split to separate the speakers, and Add subtitles for your video.