Words alone miss half the story. Get the overall sentiment of a video, its emotional tone, how exciting it is and how controversial it is, plus the sentiment toward the people and brands in it.

How it works

  1. Send video URLs in batches of up to 500.
  2. Choose sentiment on its own or with other signals.
  3. Use the signals to rank, report or flag content.

What comes back

{"type": "SENTIMENT", "result": {"overallSentiment": "mixed", "signals": {"emotionalTone": "Playful, curious, excited", "excitementLevel": "medium", "controversy": "low"}}}

What teams build with it

  • Social listening on video: what people say and how it feels.
  • Brand sentiment and context reporting: how a brand shows up, not just whether it does.
  • Flagging controversial content before it is monetized.

Frequently asked questions

Is sentiment per video or per entity?

Both: an overall read for the video, plus sentiment for the people and entities in it.

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.