This listing was selected and written with the help of artificial intelligence, and manually reviewed by a human editor before publication.
Describewright adds audio description to videos hosted on a WordPress site, for the benefit of blind and low-vision viewers who can hear what is said in a video but miss what is only shown on screen: a slide, a name on a lower third, a chart, someone walking into frame.
This matters because WCAG 2.1 Level AA requires audio description for prerecorded video under success criterion 1.2.5, a requirement behind several accessibility laws that apply to public bodies, schools and other organizations. The plugin is built for site owners who need to meet that standard without hiring a studio for each video, while keeping a human check on every description before it goes live.
The plugin relies on WordPress’s own AI client, available from WordPress 7.0 onward, rather than connecting to an AI service on its own.
You set up an AI provider once under Settings, Connectors, with an API key for that provider, and Describewright asks WordPress to send the request through that connection; whichever provider you choose then handles the request and may charge you for its own usage.
For each video, Describewright finds the pauses in the speech, picks still frames at the moments of scene change, and sends them to a vision-capable AI model, which writes a draft description for each pause. Drafts appear in an editor inside Media, Audio description, next to the playing video, where you review, adjust and save them before they go out to visitors through a player button.
Powered by Fabio ChatBot
Is Describewright yours?
Tell readers what is limited and unlimited in the free version, and what the AI does in your plugin. Help them decide faster and reach the people already looking at it. In return, you answer their questions.
Key Features of Describewright
Automatic pause detection. Describewright listens to a video’s sound track to locate the gaps between spoken sentences, which is where a description can be inserted without talking over anyone. When captions from Captionwright are available, it can use those instead to find pauses more precisely.Possible use A college publishing recorded lectures could rely on this to locate natural gaps in a professor’s narration without manually scrubbing through each recording.
Scene-aware frame selection. The plugin detects scene changes in the video and sends the still frames from those moments to the AI model, so descriptions are based on what is actually shown rather than an arbitrary frame.Possible use A museum sharing a gallery walkthrough video could use this so that descriptions line up with each new exhibit or artwork as it appears on screen.
Pause-sized description drafts. Each draft description is given a word budget based on the length of the pause it needs to fit into, and the editor flags any description that runs too long to be read in time.Possible use A nonprofit volunteer editing descriptions for a fundraising video could spot an overlong draft immediately and trim it before publishing.
On-screen text and context notes. The AI is instructed to read on-screen text such as titles, names, signs and numbers word for word, and you can add short notes once, like a speaker’s name and role, for it to reuse throughout the video.Possible use A city government describing a council meeting could add the mayor’s name once and have it carried into every relevant description automatically.
Review editor with live preview. Every draft description is shown alongside the frame it refers to, with editable start and end times and a read-aloud preview, letting you check accuracy before saving.Possible use An accessibility coordinator at a university might use the editor to confirm each description matches the slide on screen before approving the batch.
Accessible player button. Published videos get an audio description button inside the core Video block or the video shortcode, with a language picker when more than one language is available; the button works with keyboards and screen readers.Possible use A course platform could let learners turn the audio description button on or off as they watch a lesson, depending on what they need.
Synchronized transcript and caption files. Descriptions can appear as on-screen text beneath the video while they are read, and a described transcript block or shortcode lists speech and descriptions together in time order, with clickable timestamps. Standard WebVTT description files are also saved to the uploads folder for you to keep.Possible use A library archiving lecture recordings could offer a full described transcript so visitors who prefer reading can follow along without playing the video.
Works alongside Captionwright. When Captionwright has generated captions for a video, Describewright can use the timing and wording of those captions to place pauses more precisely and give the AI more context about what is being said nearby.Possible use A membership site already using Captionwright for captions could add Describewright to the same videos without setting up separate timing information.
How Describewright Compares
Swipe sideways to see every column.
| Plugin | DescribewrightThis page | Captionwright | AltScribe |
|---|---|---|---|
| AI providers | Multi-Provider | OpenAI / ChatGPT | Multi-Provider |
| Free access | Not specified | No API Key Required, Unlimited Free Version | Bring Your Own API Key |
| Works with | Not specified | Not specified | Not specified |
| Ease of use | No-Code Setup | Not specified | Not specified |
| Main focus | AI-generated audio description for video | Browser-based captions and transcripts for video | Bulk alt text generation for images |
| What sets it apart | It focuses on narrating what is shown on screen during speech pauses, using scene-aware frames and pause-fitted drafts for blind viewers. | It runs Whisper speech recognition locally in the browser with no API key or upload, unlike the other plugins that depend on connected AI providers. | It targets missing alt text across the Media Library and embedded page images, using vision models with multi-provider fallback for bulk runs. |
| Would suit you if… | you need to meet WCAG audio description requirements for prerecorded video with a human review step before publishing | you want captions, subtitles, and transcripts generated without an account, API key, or external server upload | you need to fill in missing image alt text across a large media library quickly using your own AI provider |
Good to Know About Describewright
According to the plugin’s FAQ, Describewright requires WordPress 7.0 or later because it depends on the WordPress AI client rather than connecting to a provider directly. The developer notes that AI-written descriptions are a first draft only and should be reviewed against the video before saving, since no automated description is guaranteed to convey every important visual detail.
Description files are stored as WebVTT files in the uploads folder, one per video and language, and removing the plugin deletes these files along with its settings and records, though the original videos are never altered.
If a video’s sound track cannot be decoded, the plugin stops rather than guessing at pause timing, and suggests adding captions, trying another browser, or describing without speech timing.
Download Describewright on WordPress.org