Open Source Real-time Speech Transcription in Your Browser

Whisper Transcribe is a free, open source extension that uses AI to convert speech to text entirely locally. Your audio never leaves your device.

Whisper Transcribe Demo

Key Features

Powerful speech transcription with privacy at its core

Real-time Transcription

Convert speech to text in real-time as you speak, with minimal delay.

100% Local Processing

All audio processing happens directly in your browser. No data is sent to external servers.

Multiple Languages

Support for multiple languages with our multilingual Whisper models.

WebGPU Acceleration

Utilizes your GPU for faster processing when available through WebGPU.

Easy-to-use Interface

Simple, intuitive controls for starting, stopping, and copying transcriptions.

Copy to Clipboard

Easily copy your transcriptions to clipboard with a single click.

How It Works

Powered by transformers.js and Whisper AI

1

Capture Audio

The extension uses the Web Audio API to capture audio from your microphone.

2

Process Locally

Audio is processed entirely in your browser using transformers.js and the Whisper AI model.

3

Display Results

The transcribed text is displayed in real-time and can be copied to your clipboard.

Powered by WebGPU

Utilizes your GPU for faster processing when available, falling back to CPU when needed.

CPU

Always available

GPU

For faster processing

Model Options

Choose the right Whisper model for your needs

Tiny

39M parameters

Fastest option with good accuracy for simple transcription tasks.

SpeedVery Fast
AccuracyGood
Size~80MB

Small

244M parameters

Balanced option with better accuracy and reasonable speed.

SpeedFast
AccuracyVery Good
Size~500MB

Base

74M parameters

More accurate than Tiny with slightly slower performance.

SpeedMedium
AccuracyBetter
Size~150MB

Privacy First

Your audio never leaves your device. All processing happens locally in your browser.

  • No audio data is sent to external servers
  • No transcriptions are stored or transmitted
  • Only downloads model files when you select them
  • Open source code for transparency

Privacy Comparison

Whisper Transcribe
100% Local Processing
Cloud Services
Sends Audio to Servers

Frequently Asked Questions

Common questions about Whisper Transcribe

How accurate is the transcription?

Accuracy depends on the model you choose, audio quality, and background noise. The larger models (Small, Base) provide better accuracy than the Tiny model. For best results, use a good microphone and minimize background noise.

Does this work offline?

Yes, once you've downloaded the model files, the extension works completely offline. The model files are cached in your browser, so you only need to download them once.

How much storage space do the models require?

The Tiny models are approximately 80MB, Base models are around 150MB, and Small models are about 500MB. The models are downloaded and stored in your browser's cache.

Which browsers are supported?

The extension is designed for Chrome and Chromium-based browsers (like Edge and Brave). For optimal performance, browsers with WebGPU support are recommended, but the extension will fall back to CPU processing if WebGPU is not available.

Can I use this for live meetings or lectures?

Yes, the extension is designed for real-time transcription. You can use it during meetings, lectures, or any other scenario where you need to transcribe speech as it happens.

Ready to Try Whisper Transcribe?

Install the extension and start transcribing speech in seconds