How to turn a voice message into text in Telegram

Potok transcribes Russian speech; this site has been translated into English for convenience.

A voice message or video note in Telegram has a transcription button that turns it into text right in the chat: a few times a week without Premium, with no limit with Premium. For files, links and speaker labels, you can forward the recording to Potok's (potok.now) bot @potok_now_bot: a short transcript comes back as text, a long one as a file, with 60 minutes per month for free.

A voice message in Telegram can be transcribed in two ways: with the messenger's own button, a few times a week without Premium and with no limit with Premium, or by forwarding it to a bot if you need a file, speaker labels or other formats.

How a Telegram voice message turns into text

Transcription takes a couple of steps:

1. Open Potok's Telegram bot @potok_now_bot and press Start (the /start command).
2. Forward the voice message to the bot in a private chat, straight from the chat where you have it.
3. Wait for the reply: a short recording comes back as text right in the chat, a long one as a file.
4. Under the text, the buttons Summary, Tasks and deadlines, Ask and Other format appear; press the one you need.
5. If the recording has several voices and you need speaker labels or timestamps, turn this on in advance in the bot settings (/settings).

The same file can also be uploaded in the web version of Potok's bot; it is more convenient for transcribing a recording of a whole meeting rather than one short voice message.

What the bot sends after transcription

After transcription the chat gets not raw text but text that has already been proofread:

• the text itself, in the chat (a short recording) or as a TXT or DOCX file (a long one);
• punctuation and capital letters in place and obvious recognition errors fixed, without shortening or retelling;
• timestamps in the [mm:ss] format and speaker labels Speaker 1, Speaker 2, if they were turned on in the settings;
• the buttons Summary, Tasks and deadlines, Ask, Other format and up to five buttons of your own with a ready request to the text;
• ready export formats: TXT, DOCX, Markdown, PDF, SRT, VTT; the recording stays in My recordings and can be opened again at any time.

When the bot will not fit a voice message

Potok's bot will not fit if the voice message is not in Russian: it recognizes only Russian speech and has no translation into another language. Do not expect the speakers' real names in the reply: the bot tells them apart only as Speaker 1, Speaker 2. Transcription during a call is not available either: the bot works with a finished recording and does not join the conversation itself. It will not accept an Instagram or TikTok link, only voice messages, files, YouTube, VK Video and RuTube. And the bot does not work without the internet: it is cloud-based, and recognition runs on Potok's servers in Russia.

If after this section you need a broader view, the comparison of transcription services shows the difference between the bot, apps and manual transcription.

Frequently asked questions

Primary sources: core.telegram.org · www.ixbt.com · developers.sber.ru · t.me