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

# AI Video Enhancement

AI Video Enhancements is a Disco AI workflow that automatically generates transcripts, summaries, chapter markers, and subtitles for eligible video and audio assets in your media library.

<img src="https://mintcdn.com/disco-e32b1565/jUJZ3DBb4c0FXLi6/images/AI-Video-Enhancement.png?fit=max&auto=format&n=jUJZ3DBb4c0FXLi6&q=85&s=fd1cb7d1711cc18893cb9568219535df" alt="AI Video Enhancement" width="2240" height="1120" data-path="images/AI-Video-Enhancement.png" />

## Overview

AI Video Enhancements processes video and audio assets to produce structured learning content without manual effort. For video assets, the workflow generates a transcript, a summary, chapter markers, and an auto-generated subtitle track. For audio assets, it generates a transcript, summary, and chapters. Subtitles are not available for audio. The feature is especially useful when a media asset appears across multiple lessons, pages, or event recordings, because the AI-generated content supports every place that asset is reused.

## Open a Media Asset

The most direct path to AI Video Enhancements is through the media library.

1. Go to **Admin** → **Content** → **Media Assets**.
2. Use the search bar or file-type filters to locate the target asset.
3. Open the row menu for the asset and click **Details**.

The **Details** drawer opens full-screen and displays the asset name, a media player, and AI enhancement controls. This drawer is where you generate, review, edit, and publish AI enhancement output for eligible video and audio assets.

<Info>
  The **Details** action appears for video, audio, and PDF assets. For AI Video Enhancements, only video and audio assets are eligible.
</Info>

## Generate AI Enhancements

If no AI assets exist yet for an eligible media asset, the player displays a generation button in place of any transcript content.

1. Open the **Details** drawer for a video or audio asset.
2. Click **Generate AI Transcript and Summary**.
3. Wait for processing to complete. The drawer displays a spinner and a Processing state while the job runs. Processing may take a few minutes. You can check back later or refresh the page.

Once generation succeeds, the player expands to show three tabs: **Overview**, **Transcript**, and **Subtitles**.

<Tip>
  We recommend generating AI enhancements before publishing a media asset to members. Reviewing the transcript and summary in draft state gives you a chance to catch errors and edit chapters before anything surfaces to your Academy.
</Tip>

## Review Enhancement Output

Each tab surfaces a different part of the AI-generated content.

**Overview tab:** The main editable area. Displays the AI-generated **Summary** and **Chapters**, along with edit controls for admins with the right permissions.

**Transcript tab:** The full transcript reading experience. The transcript is linked to media playback, so clicking a passage navigates to the corresponding timestamp. This tab is for reading and navigation, not direct manual editing.

**Subtitles tab:** Available for video assets. Displays subtitle controls and the auto-generated subtitle output. If no subtitle track existed before generation, the system creates an auto-generated track shell so subtitles have a place to land once processing completes.

## Edit Summary and Chapters

You can update the AI-generated summary and chapter markers at any time from the **Overview** tab.

1. Open the **Overview** tab in the **Details** drawer.
2. Click the edit icon.
3. Update the Summary field or the Chapters field as needed.
4. Click **Save Changes**.

Saving writes your updates back to the transcription record. Edits apply to the asset everywhere it is used across your Academy.

### Chapter formatting rules

The chapter editor validates your input before saving. Each chapter must follow this format:

* One chapter per line
* A timestamp followed by a chapter title (for example: 01:30 Introduction to Pricing)
* Timestamps in MM:SS or HH:MM:SS format
* Chapters listed in ascending order
* At least two chapters total

## Publish or Hide AI Enhancements

The transcription record has a visibility toggle that controls whether members can see the AI-generated content.

1. Open the **Details** drawer for the asset.
2. Click the eye icon to toggle visibility between **Published** and **Draft**.

When set to **Published**, the transcript, summary, chapters, and subtitles are visible to members. When set to **Draft**, the content is hidden from members and only accessible to admins with media-transcription permissions.

## Understand Eligibility and Access

Several conditions determine whether AI Video Enhancements appear for a given asset.

**Supported media types:** The feature runs on video and audio assets only. Images and documents are not eligible.

**Synthetic audio:** Assets generated from synthetic audio are excluded from transcription generation.

**Entitlements:** The generation button checks for an active entitlement based on media type. Video assets require the ai\_video\_transcription entitlement; audio assets require ai\_audio\_transcription.

**Feature flag:** AI Video Enhancements must be enabled for your workspace. If the feature is not active, the enhancement tabs and generation controls will not appear even for eligible assets.

**Permissions:** Only admins with media-transcription management permissions can see the generation button and edit enhancement output.

## Handle Errors and Retries

AI Video Enhancements surfaces clear feedback when something goes wrong.

**Generic processing failure:** If generation fails, the drawer offers a re-generation action. Click it to retry without rebuilding the asset from scratch.

**No audio detected:** If the media file has no detectable audio, the drawer displays No audio detected and generation is blocked. This is expected behavior for silent video files.

## FAQ

<AccordionGroup>
  <Accordion title="Does AI Video Enhancements work on audio-only files?">
    Yes. The workflow runs on both video and audio assets. Audio assets generate a transcript, summary, and chapter markers. Subtitles are not available for audio.
  </Accordion>

  <Accordion title="Can I edit the transcript directly?">
    No. The **Transcript** tab is read-only. You can edit the **Summary** and **Chapters** fields on the **Overview** tab. If you need to correct transcript text, contact Disco support.
  </Accordion>

  <Accordion title="What happens if I edit the summary or chapters?">
    Your changes update the transcription record for that asset. Because the record is shared, the updates appear everywhere the asset is used across your Academy.
  </Accordion>

  <Accordion title="Can I re-run generation after editing?">
    If generation fails, you can retry using the re-generation action in the drawer. Re-running on a successfully processed asset is not available. Edit the existing output instead.
  </Accordion>

  <Accordion title="Why don't I see the Generate AI Transcript and Summary button?">
    Several things can prevent the button from appearing: the asset is not a video or audio file, your workspace does not have the AI Video Enhancements feature enabled, you do not have the required entitlement, or you lack the media-transcription management permission. Check with your workspace owner if you expect access but do not see the option.
  </Accordion>

  <Accordion title="What does &#x22;No audio detected&#x22; mean?">
    It means the video file contains no detectable audio track. Generation cannot proceed. This is expected for silent video files and is not a product error.
  </Accordion>
</AccordionGroup>