Deepfake detectionVoice: speakers and sound

AudioSeal

An imperceptible mark in synthetic speech plus a fast detector that finds it even inside a fragment of a long recording. The detector errs in both directions: a hit is a reason for a human to check, not proof.

Developer
Meta, USA
First release
Feb 2024
Latest release
Dec 2024
Sizes
a watermark generator and detector, 16-bit message
License
Commercial use allowedMIT, weights included
Running
On your own serverAlso runs without a GPU
Industries
Media and production, Security, Customer support, Marketing and content
Deploy on your server Hugging Face GitHub

What it does

  • Marking speech synthesized by your service
  • Finding your own mark in third-party publications
  • Checking whether synthesis was mixed into a call recording
  • Tracing the source of audio inside the company

Where it is used

Audio and video productionVoice services and call centersInformation securityMarketing and brand control

Hardware requirements

LaptopLaptop or regular PC, up to 8 GB of VRAM — smaller versions
fits
1 GPUOne GPU with 16–80 GB — mid-size versions
no versions
ClusterServer with several GPUs — flagship versions
no versions

Versions

  1. AudioSeal 0.2 с потоковым режимом
  2. Открытие весов под MIT
  3. AudioSeal 0.1.2 (генератор и детектор)

How to run it

On your own serverWeights are downloaded from Hugging Face and served with vLLM (load and API) or llama.cpp (modest hardware). The model then runs in a closed network with no per-request fees.Hugging Face
How much hardware you needCalculate the VRAM for the model size, context length and number of concurrent requests.Open the hardware calculator

I can set this up end to end: pick the model size, deploy it on your server and connect it to your systems.

Frequently asked questions

Can AudioSeal be used in a commercial project?

Yes. License: MIT, weights included. It allows commercial use, but it is still worth having a lawyer review the license before launch.

What hardware does AudioSeal need?

At minimum: Laptop or regular PC, up to 8 GB of VRAM — smaller versions. Some versions also run on an ordinary CPU, without a GPU. You can calculate the exact VRAM for your model size and context in the hardware calculator.

Does AudioSeal support Russian?

Language does not matter for this model: it does not work with text.

Where can I download AudioSeal and what does it cost?

The AudioSeal weights are open and free to download. You only pay for the hardware it runs on and for the setup. Source links are at the bottom of this page.

How I deploy it for clients

  1. SelectionI pick the model size for your task and hardware and test it on your examples.
  2. DeploymentI deploy it on your server or in a closed network and provide an API.
  3. Fine-tuningI fine-tune it on your data (LoRA) or connect a knowledge base — whichever is cheaper for the task.
  4. IntegrationI connect it to your CRM, ERP, bot, website or team chat and set up monitoring.

Similar models

Deepfake detectionVideoSealMeta · USACommercial use allowed

A watermark for video and images that survives re-encoding and cropping. The detector errs in both directions: a missing mark does not prove a forgery, and finding one is a reason for a human to check.

DetailsDeepfake detectionWatermark Anything (WAM)Meta · USACommercial use allowed

An image watermark that can be applied to individual regions: the model shows which part of the image is marked. It errs in both directions - a human reviews the result.

DetailsDeepfake detectionTrustMarkAdobe Research and University of Surrey · USACommercial use allowed

An image watermark for arbitrary resolutions built for the Content Authenticity Initiative: it can both apply a mark and remove one. The detector errs in both directions - a human reviews the output.

Details

Source: github.com/facebookresearch/audioseal. Data checked against the model card on 22 Sep 2026. Have a lawyer review the license before commercial launch.