Resemble AI
If you came here for voice cloning, the product you remember has been withdrawn: Resemble now sells the opposite service, catching cloned voices and faces, and the pricing and API reflect that. As a detection vendor it is a credible shortlist entry next to Pindrop and Reality Defender, with per-second metering that is easy to trial and an on-prem option for regulated buyers.
Pros
- One REST API covers audio, video and image detection, watermarking and speaker identity
- Pay-per-use Flex tier lets you measure accuracy on your own data before committing to seats
- Containerised on-premises install for buyers who cannot send media to a cloud
- Published per-unit rates and a public benchmarks page, which most detection vendors keep behind sales
Cons
- Voice generation has been discontinued as a commercial product; existing TTS integrations need a new vendor
- Per-second detection pricing adds up quickly on long recordings or video
- Seat plans jump straight from $0 to $350 a month with nothing in between
- No headquarters address on the terms page, only Delaware as governing law