In this guide
A video transcript is a text version of a video's spoken content and relevant sounds. A descriptive transcript also conveys important visual information. Review it against the video, and provide timed captions separately when viewers need them.
A video transcript is a text version of what happens in a video, including the dialogue and relevant sounds. Depending on its purpose, it may also describe important visual information, identify speakers and include timestamps. People can read it independently of the video, search it for a phrase or use it as a starting point for captions.
For accessibility, the detail matters: a transcript that only records dialogue can miss a demonstration shown silently on screen. The W3C's transcript guidance distinguishes a basic transcript from a descriptive transcript that also communicates visual information.
What does a video transcript look like?
Here is a short, invented example from a tutorial. Timestamps are useful for navigation, but the words and descriptions remain understandable without playing the video.
[00:00] Visual: The presenter holds up a cable with a damaged connector.
Presenter: “This cable was causing the signal to cut out.”
[00:04] [Click] Visual: The presenter plugs in a replacement cable.
Presenter: “The signal is steady now.”
The sound cue matters because it marks the connection; the visual descriptions explain what the presenter shows rather than says. For an interview, consistent speaker labels may be more important than describing an unchanged background. Include on-screen text and actions when they carry meaning that the dialogue does not convey.
Transcript vs. captions vs. subtitles
| Format | How it is used | Key difference |
|---|---|---|
| Transcript | Read as a document or on a page alongside the video. | Can include dialogue, meaningful sounds and visual descriptions; timestamps are optional. |
| Captions | Displayed in sync with video playback. | Provide timed speech and relevant sound information; their presentation must be checked in the player. |
| Subtitles | Displayed in sync with playback, often to translate dialogue. | May omit sound cues and speaker information needed by someone who cannot hear the audio. |
Terms vary across platforms, so inspect the actual text and timing rather than relying on a label. A transcript alone does not replace synchronized captions, and captions alone may not communicate visual information. Some videos also need audio description. Consult the W3C's media accessibility guidance for the needs of your content and audience.
How to create a video transcript
- Choose your source and format. Use a video you have permission to process. Decide whether the reader needs a verbatim record, an edited reading version, or a descriptive transcript. Keep the original video available for checking.
- Make a first draft. Type the dialogue while pausing and replaying, or use a transcription service that accepts your file and language. If the video has a usable caption track, it can be a starting point. Check a service's storage and privacy practices before uploading sensitive material.
- Identify speakers and meaningful sounds. Use consistent names or labels; do not guess someone's identity. Add cues such as [door slams] when they help explain the action. Mark inaudible words instead of inventing them.
- Watch for visual information. Add descriptions of necessary actions, diagrams and on-screen text that are not explained in the audio. Make them concise and place them near the relevant dialogue.
- Review against the original. Check names, numbers, technical terms, speaker changes and negations. Automatic drafts can miss words or assign them to the wrong person. Add timestamps if they help readers find sections.
- Publish and test. Put the transcript in an accessible location near the video or link to it clearly. Check that its text can be selected and read on a phone. Prepare and check timed captions separately.
For specific transcription methods and privacy checks, see how to transcribe audio to text. If you need text synchronized with playback, follow how to add subtitles to a video.
Can you view a transcript on YouTube?
For the platform-specific viewer and creator steps, see how to get a YouTube transcript or download your captions.
For a YouTube video with captions, open its description and look for Show transcript. You can select a line to jump to that point in the video. Availability and controls vary, and not every video has a transcript. YouTube explains the current steps in its transcript instructions.
Automatic captions and transcripts can contain errors. If you are preparing your own transcript, compare any available draft with the original before publishing, and verify that you have the necessary rights to reuse it.
What if the audio is hard to hear?
Listen to the original in short sections and use headphones where helpful. If you have permission to prepare an audio copy, the audio trimmer can create a short excerpt from a supported local MP3 or WAV file; check its current limits and keep the excerpt's start time so it can be mapped back to the video. It edits audio and does not transcribe it.
When a competing sound masks dialogue, an isolation estimate may help you listen, but it can omit or alter speech. Compare it with the original and leave uncertain words marked. Our guide to audio isolation versus noise reduction explains the tradeoff.
Common questions
Is a video transcript the same as a script?
No. A script is usually written before recording; a transcript represents the finished recording. A speaker may improvise, and the final video may contain sounds or visual details absent from the script.
Does a transcript need timestamps?
Not always. Timestamps can help readers locate passages, especially in a long video. If the goal is timed captions, use a caption format and check the timing against playback.
Can I use an automatic transcript without editing it?
A draft can save time, but review it against the original. Accents, overlapping voices, background sound and specialized terms can cause errors, and automated speech text may omit important visual information.