Voice Note App

This application requires WebGPU.

Please ensure WebGPU is enabled in your browser settings. In Chrome, you can enable it by going to chrome://flags/#enable-unsafe-webgpu and enabling the "Unsafe WebGPU" flag.

Initializing AI Model...
Preparing your voice transcription experience
audio2text
.email
sign in

Voice Note

0 words 0 characters

Getting Started & How to Use

How do I transcribe my voice notes with audio2text.email?
Transcribing your voice is simple with our free tool:
  1. Select Your Language: Choose your desired language from the available options.
  2. Start Recording: Click the microphone button in the text editor. This will begin capturing audio from your microphone.
  3. Stop Recording: Click the microphone button again to stop.
  4. See the Magic: Our app will then transcribe the recorded audio in chunks (usually 30 seconds or less). The transcribed text will appear in the editor, replacing the current cursor selection.
How does the voice to text transcription actually work?
Our app uses advanced technology right in your web browser:
  • Audio Capture: When you record, the app listens for natural pauses or records up to 30 seconds of continuous speech.
  • Local Processing: This audio chunk is then processed directly on your device using a downloaded AI model (powered by transformers.js and running via WebGPU if your browser and device support it).
  • Privacy Focused: Because the transcription happens locally in your browser, your audio data doesn't need to be sent to external servers for processing after the initial model download, enhancing your privacy.
What AI model does Voice Note use for transcription?
Our app selects the AI model based on your device to ensure optimal performance: All powered by the transformers.js library.
How accurate is the transcription compared to OpenAI's API?
The AI models we use are derived from the same powerful technology as those used by OpenAI's API. However, the models running in your browser are optimized for speed and local processing, meaning they are generally smaller. While they provide excellent and fast free transcriptions, the largest models accessible via OpenAI's paid API might offer slightly higher accuracy for very complex audio. Our goal is to provide the best possible accuracy for a free, browser-based tool.
Is there a time limit for how long I can transcribe?
You can transcribe for extended periods, even multiple hours, provided:
  • Your computer has a GPU (Graphics Processing Unit).
  • Your web browser supports WebGPU technology.
The primary practical limitation might be the maximum amount of text a standard HTML textarea element (which our editor uses) can hold. For most voice note purposes, this should be more than sufficient.
Is this voice note transcription app really free?
Yes, our voice note transcription app is completely free to use directly in your browser.
Which web browsers and devices are supported?
Our voice-to-text app currently works best on:
  • Google Chrome: Desktop and Android versions.
We are constantly working to improve compatibility and hope that as web technologies evolve, this app will work seamlessly across all major browsers.
Is my audio data safe? What about my privacy?
Your privacy is a top priority. Here's how we handle your data:
  • Local Processing: The core transcription process happens directly within your web browser on your device. Your audio is not uploaded to our servers for transcription after the initial AI model is downloaded to your browser.
  • No Audio Storage: We do not store your voice recordings or transcribed text on our servers.
This local processing approach significantly enhances your data security and privacy.