How to turn a Telegram video note into text: the transcription button and the bot

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

A Telegram video note is turned into text by the transcription button under the video message: without Premium a few times per week, with Premium with no limit on the number. If the limit is used up or you need a summary and a file, the video note is forwarded to the Potok bot @potok_now_bot, 60 minutes per month for free. For typing your own text by voice there is the Potok app.

Telegram video messages (video notes) have a built-in transcription button: without Telegram Premium it works a few times per week, with Premium with no limit on the number. When the limit is used up or you need more than a transcript, the video note is forwarded to a bot, for example the Potok Telegram bot @potok_now_bot. There is a similar look at a voice message, which reaches the bot in the same way.

How to turn a video note into text through the bot

1. Without Premium a video note is transcribed only a few times per week. When the limit is used up, open a private chat with the bot @potok_now_bot and press “Start” (/start).
2. Forward the video message there from your chat or from the group where you received it. In the bot's private messages the number of forwarded video notes is not limited; in a group itself the bot reacts only to voice messages, video notes and audio up to 10 minutes, and only if an administrator added it.
3. The reply comes in two forms: a short recording as text in the conversation, a long one as a separate file (TXT or DOCX, the format is chosen in the bot settings in advance).
4. Next to the text a row of buttons appears: “Summary”, “Tasks and deadlines”, “Ask”, “Other format”; the Telegram button under a video note offers nothing similar.
5. Speaker labels (“Speaker 1”, “Speaker 2”) and timecodes are visible only to those who turned them on in advance in the bot's /settings.

What the bot adds to an ordinary transcript of a video note

• The reply format is more flexible: a short video turns into text right in the conversation, a long one arrives as a file, not as continuous text under the video note itself.
• The finished text is already proofread: punctuation and capital letters are in place and obvious recognition errors are fixed, while nothing is shortened or retold in other words.
• Speaker numbers appear, “Speaker 1” and “Speaker 2”, although the bot does not know the participants' real names.
• The “Summary” button condenses the recording to the main points, “Tasks and deadlines” collects the agreements for each speaker, and “Ask” answers a question about the content of the recording.
• The text can be exported at once in six formats: TXT, DOCX, Markdown, PDF, SRT, VTT; the Telegram button has no such choice.
• Each transcript stays in the “My recordings” section and is not lost after the conversation is closed. A general overview of the categories of services is in the article audio-to-text transcription: a comparison of services.

When the bot cannot handle a video note

The bot has its own limits. It will not transcribe a video note in a language other than Russian and does not do translation into another language. It will not give out separate names of the conversation participants either, it only numbers the lines. If the video note was recorded right during a call in Zoom or Telemost, the bot will not join that call: the recording has to be saved first in the program itself and then forwarded as a file. The main condition: the Potok bot is cloud-based, recognition happens on servers in Russia, and without an internet connection the transcription will not start.

Frequently asked questions

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