See what the camera sees. For every video, get the genre, setting and tone, the scenes, objects, logos and on-screen text, and a content overview: theme, subject, language, content type and keywords.

How it works

  1. Send video URLs in batches of up to 500.
  2. Choose visual analysis and content overview. Visual analysis reads storyboard frames sampled across the video.
  3. Build on the structure: scene metadata, placement checks, brand reporting.

What comes back

{"type": "STORYBOARD_ANALYSIS", "result": {"analysis": {"overall_scene_and_setting": "A cinema auditorium with red reclining seats", "overall_genre": "news, event coverage", "overall_objects": "handheld microphone, popcorn cartons, projection screen", "overall_texts": "REWIND YOUR MIND, POP CORN"}}}

What teams build with it

  • Product placement QA: confirm the product actually appeared.
  • Sponsorship measurement: which brands appear, and where.
  • Scene-level metadata for video platforms and libraries.

Frequently asked questions

Does it detect brands and logos?

Yes, logos and on-screen text appear in the analysis. Full timestamped brand and logo detection is part of Content Intelligence; tell us what you need to measure.

Is it frame by frame?

No. It analyzes storyboard frames sampled across the video, which keeps large libraries fast and affordable.

Whose models, and at what scale

Our own models, trained and curated in-house by advertising and media people, so we can tune them to your needs. Your video isn't sent to a big LLM provider.

Built for libraries and feeds: our infrastructure processes 100k+ videos per hour.

New signals added regularly. Custom models on request.

Ready to test it on your own video? Get API access.