★ My Tools

Voice to Text – Convert Telugu and English Recordings to Text

Upload a press meet or interview recording and get the text. Works for Telugu and English. You can also speak live, and translate the text below. Recordings are not uploaded.

🔒 Recording stays on your device · translation uses Google✓ Free · No sign-in

🎧 Click or drag & drop an audio or video recording

MP3, M4A, WAV, OGG, AAC, MP4, WEBM and more. The file stays on your device.
Telugu needs the Best quality. The model is downloaded once and kept in your browser.
🌐 Live typing uses the speech service of your browser (Google Chrome or Microsoft Edge). Your voice is sent to that service. Firefox does not support it. For private or long recordings, use "Upload a recording".

📝 Text

🌐 Translate this text

🌐 Translation uses Google Translate. Only the text is sent, never the audio. Do not translate confidential text.

How to Convert a Recording to Text

Record the speaker on your phone, then bring the recording here. The tool listens to it inside your browser and writes the text.

  • Click "Select Recording" and choose the audio or video file
  • Choose the language spoken in the recording
  • Choose the quality. For Telugu choose "Best". The first time, the speech model is downloaded once
  • Click "Convert to text" and wait. The text appears part by part while you wait
  • Read and correct the text, then copy it or save it as a .txt or .srt file
  • To translate, choose a language under "Translate this text" and click Translate

Everything happens inside your browser, so your document is never uploaded.

Why Use This Tool?

At a press meet, writing down every word is hard, and later listening again to write the news takes hours. With this tool you record on your phone, upload the file and get a first draft of the text within minutes.

Your recording is not sent to any server. The speech model runs inside your own browser, so private interviews stay private. Only the optional translation sends the text (not the audio) to Google Translate.

Features of Voice to Text

  • Telugu, English, Hindi, Tamil, Kannada, Malayalam, Marathi, Bengali and Urdu recordings
  • Works with MP3, M4A, WAV, OGG, MP4 and other audio or video files
  • Recording is processed on your own device
  • Text appears part by part, with time stamps you can switch off
  • Save as .txt or as a .srt subtitle file
  • Live typing with the microphone in Chrome and Edge
  • Translate the text into 16 languages
  • Stop any time; the text so far is kept

Frequently Asked Questions

Is my recording uploaded?

No. The recording is read and converted inside your browser. Only the first-time download of the speech model comes from the internet.

Why does the first conversion take longer?

The speech model (40 to 250 MB) is downloaded once. After that it is kept in your browser and starts quickly.

How accurate is it?

Clear speech with little noise gives good results. Telugu is harder than English, and names, numbers and mixed Telugu-English speech can have mistakes. Use the "Best" quality for Telugu and always read and correct the text before you publish.

How long does it take?

It depends on your device and the quality you choose. Fast and Normal are quicker than the audio length on most computers. Best can take as long as the recording or longer, and is slow on phones. Use a computer for long recordings.

How long a recording can I use?

Recordings up to about one hour work well on a computer. For longer recordings, cut the file into parts.

My phone recording is not opening.

Browsers cannot open some phone formats such as AMR. Convert the file to MP3 or M4A, or record with an app that saves as M4A or MP3.

What is the microphone option?

It types while you speak, using the speech service of Chrome or Edge. Your voice is sent to that service, so use it only for non-confidential speech.

Is it free?

Yes, completely free.

About This Tool

Voice to Text is made for journalists, editors and students who need the words of a recording in writing. It gives you a first draft that you can correct and use.

Speech recognition is never perfect. Always listen again for names, figures and quotes before publishing.