Resemble AI

Tomorrow our Head of JevRel, @0xDevShah , will be speaking at @FeatherlessAI and @Tiny_Fish's Hack and Demo night. He'll be showing how we do agent detection, kind of a must have when everyone is using agents.

Super hot topic x smart speakers = amazing event. See you there > luma.com/7x4tywqh

Resemble AI's open source models have been downloaded over 17 million times. Today we're adding our name to the Open Weights and American AI Leadership letter.

PerTh for watermarking, Resemblyzer for speaker verification, and Chatterbox and Dramabox on the generative side. Releasing these in the open is what keeps us in the research conversation rather than adjacent to it.

That research has directly led to our leading multimodal Deepfake Detection Model. You cannot reliably catch synthetic media without understanding how it's made, and we learned that by making it in public, where other researchers could push back on the work. That's why our DETECT-3B Omni model holds a top spot on benchmarks like Podonos and DFBench.

Anyone can generate any content with AI.

So how do you prove what's really yours?

Resemble Watermarker - Encode a mark and sign C2PA. EU AI Act compliant.

Available via API, try it free at app.resemble.ai/hub/watermarki…

Is this real or AI?

Even a year ago that wasn't what made you stop the scroll...but now it is.

That's why we built Resemble AI Deepfake Detector for Chrome, available here (for free) > chromewebstore.google.com/detail/resembl…

Deepfakes are everywhere, but so are we.

🎭 DramaBox - the most expressive, promptable, open source TTS model ever - is now available.

Get it on:
@huggingface - huggingface.co/spaces/Resembl…

Resemble AI Labs - app.resemble.ai/hub/labs

Woop!