Audio to Text Converter

How it works

Typing out a recording by hand takes several times as long as the recording itself. This free audio to text converter does it for you online: upload an audio or video file, leave the language on Auto detect or pick one, and the transcript is ready in under a minute for most clips. Nothing to install and no account needed.

It is built for the way India actually speaks. Most transcription tools expect one language at a time and stumble when a sentence starts in Hindi and ends in English. Here you can ask for Hinglish, which writes Hindi words in English letters the way people type them ("aaj hum paneer butter masala banayenge") and leaves English words as English. Or pick the native script: Hindi in Devanagari, Tamil in Tamil script, and so on.

Unlike our other free tools, this one cannot run inside your browser. The file is uploaded to our servers, a speech model listens to it, and the text comes back to this page. The upload is deleted as soon as the transcript is written, and the transcript itself is deleted after 24 hours.

Languages you can get the text in

Leave it on Auto detect and the tool works out which language is spoken. Next to it, choose how to write it: English letters (Hinglish, Tanglish and so on, the way people type on their phones) or the language's own script (Devanagari, Tamil script and so on). English speech comes out in English either way. You can also pick an exact language from the list.

• Hindi: Hinglish, or Devanagari
• Marathi: Minglish, or Devanagari
• Bengali: Banglish, or Bangla script
• Tamil: Tanglish, or Tamil script
• Telugu: Tenglish, or Telugu script
• Gujarati: Gujlish, or Gujarati script
• Kannada: Kanglish, or Kannada script
• Malayalam: Manglish, or Malayalam script
• Punjabi: Punglish, or Gurmukhi
• Urdu: Roman Urdu, or Urdu script
• English: plain English, translated when the speech is in another language

Already have Hindi text and only need it in English letters? The Hindi to Hinglish converter does that instantly, with no upload.

What you can upload

Audio files (MP3, WAV, M4A, AAC, OGG, FLAC, OPUS) and video files (MP4, MOV, WebM, MKV), up to 200 MB. WhatsApp voice notes, phone call recordings, Zoom and Google Meet recordings, lectures and podcast clips all work. For a video, only the sound is used. Converting an MP3 in particular? The MP3 to text converter is the same tool with MP3 tips.

Clear speech gives the best text. Loud music, several people talking at once or a very noisy room will cause mistakes, so read the result once before you use it.

Free limit

Every visitor gets 5 free minutes of audio a day without an account. A free DesiCaptions account raises that to 10 minutes a day. The count resets at midnight, Indian time.

If a file is longer than the minutes you have left, the tool transcribes the first part that fits and tells you where it stopped. The limit exists because each minute of audio costs us real money to process, and it is what lets the tool stay free.

Read, edit and download

The transcript opens as a list of lines, each with the time it was spoken. Click a line to play the recording from that point, and press Edit to fix a word or a name before you download.

Download gives you TXT, PDF, Word (.docx), SRT or CSV, or copies the text. For the text formats you choose one paragraph or several, and whether to keep the timestamps. SRT works as subtitles in YouTube, Premiere Pro or CapCut. To change an SRT into another format, use the SRT converter.

Frequently asked questions

Is this audio to text converter free?

Yes. You get 5 minutes of audio a day without signing up, and 10 minutes a day with a free account. There is no card and no trial that turns into a charge.

How do I transcribe audio to text for free?

Upload the recording above, leave the language on Auto detect, and wait for the transcript. Check names and rare words, then download it as TXT, PDF, Word or SRT, or copy it. It all happens online in your browser, with nothing to install.

Can I convert MP3 to text?

Yes. MP3 is the most common file people upload here. See the MP3 to text converter for tips on MP3 files.

Can it convert Hindi audio to text?

Yes. Choose Hindi for Devanagari text, or Hinglish for Hindi written in English letters. Sentences that mix Hindi and English are handled as they are spoken.

Do I have to choose the language?

No. Auto detect listens to the start of the recording and recognises the language. Choose English letters (such as Hinglish) or the language's own script (such as Devanagari) next to it. Pick a language from the list only when the detection gets it wrong.

Which languages are supported?

Hindi, English, Marathi, Bengali, Tamil, Telugu, Gujarati, Kannada, Malayalam, Punjabi and Urdu. Every Indian language can be written in its own script or in English letters.

Can I convert a video to text?

Yes. Upload an MP4, MOV, WebM or MKV file and the speech in it is transcribed. The picture is ignored.

Does it work with WhatsApp voice notes?

Yes. Save or share the voice note as a file (it is usually an OPUS or M4A file) and upload it here.

Is my audio uploaded?

Yes. The file is sent to our servers to be transcribed, because speech recognition at this quality cannot run in a browser. The file is deleted as soon as the transcript is written, and the transcript is deleted after 24 hours.

How accurate is it?

Clear speech from one speaker is transcribed well. Background music, crosstalk and poor recordings reduce accuracy. Names and rare words are the most common mistakes, so check them.

How long does it take?

A short clip is usually ready within half a minute after the upload finishes. Longer files take longer, and a busy period can add a short wait.

Can I get subtitles instead of plain text?

Yes. Download the result as an SRT file. For captions burned into the video with styles, use DesiCaptions itself.

Does it label different speakers?

Not yet. The text is written in order without speaker names.