---
title: "Transcribe audio and video"
description: "Turn recordings into searchable text, follow along with playback, label speakers, and download subtitles."
canonical: "https://bearly.ai/docs/features/transcripts/"
markdown: "https://bearly.ai/docs/features/transcripts.md"
---

# Transcribe audio and video

HTML: [https://bearly.ai/docs/features/transcripts/](https://bearly.ai/docs/features/transcripts/)

> Turn recordings into searchable text, follow along with playback, label speakers, and download subtitles.

**Transcribe** turns the spoken content of an audio or video file into text. The result stays connected to the recording, so you can follow along, jump to a moment, download subtitles, or continue asking questions in the same chat.

## Create a transcript



1. **Choose Transcribe**

   Type `/transcribe` in a chat, then select **Transcribe** from the command menu.

2. **Attach the recording**

   Select the plus icon beside the message box, choose **Attach files**, and add one or more audio or video files. Wait for the attachments to finish loading.

3. **Send the message**

   Bearly transcribes every attached recording. You can also add a follow-up request in the same message, such as `Transcribe these interviews, then list the decisions and open questions.`



Audio and video files can be as large as 400 MB each. Whether a particular format plays or uploads can vary by device, so use a common audio or video format if a file is rejected.

## Review the result

The transcript appears below the media player.

- **Captions** follows the current point in the recording and is useful for watching or listening along.
- **Table** shows timestamped passages. Select a passage to jump to that point in the recording.
- **Density** changes how much text Bearly groups into each passage. It does not change the words in the transcript.
- If Bearly identifies different speakers, edit the **Speaker** names shown beneath the transcript. The names are saved when you leave each field.
- When several recordings were submitted together, use the previous and next controls to move among their transcripts.
- Select **Download subtitles** to save the current transcript as an SRT subtitle file.

Bearly also attaches a text copy of a completed transcript to the conversation. This lets you ask for a summary, quotations, action items, a cleaned-up interview, or answers grounded in the recording.

> **Caution**
> Names, numbers, accents, and overlapping voices can be transcribed incorrectly. Check important passages against the recording before publishing or relying on exact wording.


## If a file cannot be transcribed

- Confirm that it is an audio or video file and that the attachment finished loading.
- If the file is larger than 400 MB, export a smaller copy or divide it into parts.
- If some files succeed and others fail, keep the completed results and submit the failed recordings again separately.
- If the transcript is empty or inaccurate, check that the recording contains audible speech. A cleaner recording or less background noise may improve another attempt.
- If an older transcript says it is no longer available, attach the original recording and run **Transcribe** again.

## Continue through the documentation

- [Documentation index](https://bearly.ai/docs.md): Browse every Bearly guide.
- [Previous: Talk with Bearly by voice](https://bearly.ai/docs/features/voice-conversations.md)
- [Next: Turn text into speech](https://bearly.ai/docs/features/text-to-speech.md)