Skip to content
Real-time AI effects for video and voice in your app

Real-time AI video effects & noise cancellation SDK

Virtual background, blur, auto-framing and beautification for cameras. Noise and room echo removal, Studio Sound and voice changing for microphones. On-device on web, iOS, Android, Windows and macOS.

Free trial · Integration help from our team

Webcam before and after Effects SDK
Noise suppression onKeyboard and fan removed

Runs on

  • Web
  • iOS
  • Android
  • Windows
  • macOS
  • Linux
  • Flutter
  • Electron

Works with

  • WebRTC
  • LiveKit
  • Zoom Video SDK
  • Cloudflare Calls
  • Custom integrations

Two SDKs, one integration experience

Use them together or on their own, with the same API style on every platform.

Video Effects SDK

Make every webcam look professional, in real time and on device.

  • Virtual background, blur and replacement
  • Auto-framing (Smart Zoom)
  • Beautification and low-light fix
  • AI color correction and LUT filters
npm i effects-sdk

Audio Effects SDK

Clear voice for calls, streams, recordings and AI voice agents.

  • AI noise suppression: speed, balanced, quality
  • Room echo removal (dereverberation)
  • Studio Sound: EQ, compressor, de-esser, AGC
  • Voice Changer: pitch, formant, chorus
npm i audio-effects-sdk

A few lines to your first effect

Install from npm, pass your Customer ID, hand the SDK a camera or microphone stream and play the processed one.

import { tsvb } from 'effects-sdk';

const sdk = new tsvb('YOUR_CUSTOMER_ID');
sdk.preload();

sdk.onReady = () => {
  sdk.useStream(cameraStream);
  sdk.run();
  sdk.setBlur(0.6);
  video.srcObject = sdk.getStream();
};
import { atsvb } from 'audio-effects-sdk';

const sdk = new atsvb('YOUR_CUSTOMER_ID');
sdk.preload();

sdk.onReady = () => {
  sdk.useStream(micStream);
  sdk.run();
  audio.srcObject = sdk.getStream();
};

Built to ship in production

Simple integration

A small, documented API with code samples. We help your team with the integration.

Cross-platform

Native SDKs for Windows, macOS, iOS, Android and Linux, plus the web, with the same API style everywhere.

Quality and speed

Models tuned for the balance of quality and real-time performance.

Efficient on device

Video runs on the GPU (DirectX 11, Metal, WebGPU, WebGL); audio runs on the CPU with WebAssembly or native code.

Simple session-based pricing

Same plans for Video and Audio SDK. One session is one SDK initialization, however long it lasts and however many effects run.

Free trial

$0

7 days with every feature on your platforms. Start in the Developer Portal, no sales call.

Startup

from $50

Pay as you go by sessions, with automatic tracking and card payment. Price depends on volume, platforms and features.

Enterprise

Custom

The lowest session price, a 1-month trial, on-premises deployment and an account manager.

Questions

Something else? Ask us below.

How do I get a trial?

Sign up in the Developer Portal and start a trial yourself: you get a Customer ID right away, then install the SDK from npm or download the native package for your platform.

How fast is it?

It depends on the frame size and format, the inference backend (GPU or CPU), the hardware and what else your app processes. On a typical laptop the Video SDK processes Full HD video at up to 50 FPS.

Can I pass GPU frames as input?

Frames are processed on the GPU: DirectX 11 on Windows, Metal on macOS and iOS, WebGPU or WebGL in browsers, with a CPU fallback. On iOS and macOS you can pass a CVPixelBuffer directly; a Metal-compatible buffer avoids extra copies. If you need another GPU input path, contact us.

How do you deliver updates?

Each release is published with release notes on the Release notes page and on GitHub. Web SDKs are on npm (effects-sdk and audio-effects-sdk); native SDKs ship as packages for each platform. Model files are versioned, so clients on the previous version keep working while you upgrade.

Do you help with integration?

Yes. We help your developers choose the best integration path for your stack and answer questions as you go. Usually we open a shared chat with your team for quick answers. If you have no developer for the integration, we can do it for a fixed price.

Can you customize the SDK?

Yes. We adapt the SDK to your requirements, usually on a time-and-materials basis.

Do you plan to add new features?

Yes. Recent releases added video noise suppression, portrait lighting, avatars, custom lower thirds and captions, and Studio Sound in the Audio SDK. Follow the release notes and tell us which features you need next.

Still have questions?

Tell us about your product and platforms. Our team will get back to you and can help with the integration.

Ready to build? Open the Developer Portal and start a free trial.

Protected by reCAPTCHA. See our privacy policy.