> ## Documentation Index
> Fetch the complete documentation index at: https://pipit.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Transcribe files

> Turn audio and video files into text with the Transcripts tab.

The **Transcripts** tab turns audio and video files into text. Transcription runs on your Mac, so your recordings stay on it.

Pipit supports WAV, MP3, M4A, AIFF, MP4, and MOV files.

## Transcribe a file

1. Open Pipit and choose the **Transcripts** tab.
2. Drag a file onto **Transcribe a file**, or click it and pick a file.
3. To label who said what, turn on **Detect speakers**.
4. Click **Transcribe**.

When it's done, the file shows the number of words (and speakers, if you detected them). Click **Open** to read the transcript. If something went wrong, click **Retry**.

If **Transcribe** is dimmed, Pipit is still loading your voice model. Wait a moment and try again.

> [!NOTE]
>
> The first time you use **Detect speakers**, Pipit downloads what it needs, so you'll need an internet connection.

## Work with a transcript

Open a transcript to:

* **Search** within it, with **Search in transcript…**.
* Get an **AI Summary**. This needs [AI cleanup](https://pipit.mintlify.site/ai-cleanup.md) set up: turn on **Clean up with AI** in **Enhance** and finish setting up your provider. The transcript's text is sent to the AI provider you picked.
* Export it. Every transcript exports as text. If you used **Detect speakers**, you can also choose **Export as SRT**, **Export as VTT** (subtitle files), or **Export as JSON**.
* **Delete** it.

**AI Summary**, export, and **Delete** are in the **⋯** menu at the top of the transcript.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.