> For the complete documentation index, see [llms.txt](https://evidano.gitbook.io/ailyze-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://evidano.gitbook.io/ailyze-docs/prepare-research-data/transcribe-audio-and-video.md).

# Transcribe Audio & Video

Turn audio and video recordings into editable, speaker-labeled transcripts you can check, correct, download, share, or send straight into analysis.

This page shows you how to turn interview, focus-group, or meeting recordings into [transcripts](/ailyze-docs/reference/glossary.md#transcript) you can check, correct, download, and analyze. At the end you will have a timestamped, speaker-labeled transcript for each recording. Every control on these screens is listed in [Transcription Settings](/ailyze-docs/reference/settings/transcription.md).

## Before you start

* Your recordings as audio or video files, or as links, cloud files, or Zoom and Teams recordings. See [Supported File Types](/ailyze-docs/reference/supported-file-types.md).
* Your free audio allowance is used first; anything beyond it is charged only after you confirm. See [Plans, Limits & Credits](/ailyze-docs/reference/plans-limits-and-credits.md).

## Step 1: Add your recordings

1. From your account page, click **Transcribe Audio/Video**.
2. On **Audio/Video Transcript Creation**, drop your files on **Local Files**, or pick a cloud tab: **Google Drive**, **Dropbox**, **OneDrive**, **Zoom**, **Teams**, or **From URL**.

You should see a tile with a green tick for each file and a line counting the uploaded files. The settings open below the files.

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2Fgit-blob-fe707626319e264bae614dc43f14b6b43feebc11%2Fsettings-transcription-01-sources.png?alt=media" alt="Audio/Video Transcript Creation page with seven source tabs above the Local Files drop zone"><figcaption><p>Files from every tab land in the same list, so one project can mix sources.</p></figcaption></figure>

## Step 2: Check the cost and name the project

Read the blue banner above the settings: it states how long the batch is, how much free time it uses, and what the rest costs. Optionally click **Project Name (optional)** to name the folder the transcripts will go in.

With several files, leave **Same for all files** to set everything once, or choose **Custom per file**.

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2Fgit-blob-7225f324109df2d302ccc6dc584341800836d779%2Fsettings-transcription-03-batch.png?alt=media" alt="Toolbar with Project Name, Use previous settings, Same for all files and the view icons above the blue cost banner"><figcaption><p>Use previous settings copies the language, dictionary, speaker and redaction choices from an earlier transcription.</p></figcaption></figure>

## Step 3: Choose the language

Click **Language of Audio** and pick the spoken language. Type to search, and pick the most specific entry, for example **British English** rather than **Global English**. For a recording that switches language, pick "2 or more languages in 1 audio (auto-detect)" at the top.

You should see the language in the field and the red "Please fix: Language is required" note gone.

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2Fgit-blob-7cb56660c9fc92789e287c13d02594ef0ce29265%2Fsettings-transcription-08-language.png?alt=media" alt="Language of Audio list open on the Best Accuracy group with the auto-detect entry first"><figcaption><p>Best Accuracy languages support every option below; Good Accuracy languages lock some of them.</p></figcaption></figure>

Optionally, choose **Translate To (optional)** to get a translated transcript as well.

## Step 4: Set up the speakers

1. Click **Optional Settings**, then **Speaker Labels**.
2. Set **Number of Speakers** if you know it; leave **Auto-detected** otherwise.
3. Under **Choose speaker identification**, pick **Role** for interviews, and fill in the roles (for example Interviewer and Participant) with a short description of each.

You should see one card per speaker under **Known values**.

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2Fgit-blob-e70c31d0c18a676b423db14772bbd764cc0bce88%2Fsettings-transcription-11-speaker-role.png?alt=media" alt="Speaker Labels with Role selected and two speaker cards for Interviewer and Participant"><figcaption><p>A description such as "asks the questions" helps Evidano tell similar voices apart.</p></figcaption></figure>

## Step 5: Add your names and terms

Under **Improve Transcription Accuracy**, type the names, places, and jargon in your recordings into **Custom Dictionary**, pressing Enter after each. Use **Custom Spelling/Formatting** to force a written form, such as "sequel" written as "SQL".

The other groups are optional: **AI Analysis** adds chapters, highlights, entities, sentiment, and topics to the editor; **Redact Sensitive Information** removes personal details; **Trim Audio** transcribes only part of a recording.

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2Fgit-blob-50afb39e898a3b233b7004b4cbe572feb9d47fe7%2Fsettings-transcription-13-accuracy.png?alt=media" alt="Improve Transcription Accuracy with a spelling rule, two place names in Custom Dictionary and the filler-words box"><figcaption><p>Custom Dictionary is the place for names: it fixes both spelling and capitals.</p></figcaption></figure>

## Step 6: Submit

Click **Process All Transcriptions**. If anything is missing, the red note above the button names it.

{% hint style="warning" %}
**Costs money:** When the batch goes beyond your free audio, a **Create project** dialog shows the charge and your balance before and after. Nothing is charged until you click **Create project** there. See [Plans, Limits & Credits](/ailyze-docs/reference/plans-limits-and-credits.md).
{% endhint %}

You should see "Analysis Starting", then the transcript list. You get an email when each folder is done.

## Step 7: Track progress

On the **Transcribe Audio/Video** list, each project is a folder with its files underneath. The status pill moves through **Validating**, **Processing**, **Initial transcript**, **Enhancing with AI**, **Transcript ready**, and **AI reviewed**; the list refreshes itself.

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2Fgit-blob-998c532ae28ca3a3fff3e04e2aead7d267c1e6f1%2Fsettings-transcription-19-list.png?alt=media" alt="Transcribe Audio/Video list with folder rows, Owner / Role, Shared with, Status, Language, Files and Date Created columns"><figcaption><p>Click a file row to open its transcript; click a folder row to show or hide its files.</p></figcaption></figure>

## Step 8: Check and correct the transcript

Open a file. The transcript shows one paragraph per speaker turn with a timestamp, and the player sits at the bottom.

1. Click **AI Edits** to see the corrections the AI review made, and **Undo** any you disagree with.
2. Click **Accuracy** and work through **Words to check**: click one to hear it.
3. Click a speaker name to open **Rename speaker**, type the real name, and choose **Rename all**.
4. Click **Edit text** (the pencil) to fix wording, then click away or **Save**.

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2Fgit-blob-c93f04fe1587ac99f9be19ddd825ccf276ecc845%2Fsettings-transcription-27-editor.png?alt=media" alt="Transcript editor with the toolbar, speaker paragraphs with timestamps and the player at the bottom"><figcaption><p>Click any word for Play from here, Edit text from here, a clip, or a new speaker.</p></figcaption></figure>

## Step 9: Download, share, or analyze

* **Download transcript** (in the editor toolbar or its **More actions** menu, or the download icon on the list) offers **DOCX**, **XLSX**, **SRT**, **VTT**, and **TXT**. Tick **Include timestamps** first if you want them.
* **Share** sends a link by email: **Add as collaborator** gives a paid Evidano user edit access, and **Share through link** gives anyone a view-only link.
* **Analyze transcript** opens document analysis with the transcript loaded; on the list, **Analyze transcripts** offers all your finished transcripts. See [Analyze Interview Transcripts](/ailyze-docs/common-research-workflows/analyze-interview-transcripts.md).

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2Fgit-blob-54fadfc4b4e13bd376e93d29176fb312f3fd3b57%2Fsettings-transcription-23-download.png?alt=media" alt="Download transcript dialog with Include timestamps and the DOCX, XLSX, SRT, VTT and TXT cards"><figcaption><p>Clicking a format downloads it straight away; there is no separate download button.</p></figcaption></figure>

<details>

<summary>Optional: common mistakes</summary>

* **Wrong language picked:** the transcript comes back garbled. Upload the file again with the right language; nothing changes the language after submitting.
* **Generic labels for an interview:** you get Speaker A and B. Fix it afterwards with **Fix speaker identification** in the editor's **More actions** menu.
* **Names in Custom Vocabulary:** use **Custom Dictionary** for names and brands; Custom Vocabulary takes spoken forms only.

</details>

## Next steps

* [Analyze Interview Transcripts](/ailyze-docs/common-research-workflows/analyze-interview-transcripts.md) — find themes across your new transcripts.
* [Transcription Settings](/ailyze-docs/reference/settings/transcription.md) — every option, dialog, and editor tool in detail.
* [Transcription Issues](/ailyze-docs/troubleshooting/transcription-issues.md) — wrong speakers, locked editing, or charges.