PDF Alt Text Tagger
A PDF alt text tagger finds every embedded image in a PDF and lets you attach a written description to each one, embedded as a real accessibility tag screen readers can announce.
Drop a PDF here to find its images
How alt text tagging works
Upload a PDF and this tool scans every page for embedded images — logos, charts, screenshots, photos, anything drawn as an image rather than real text. For each one it finds, you write a short description of what the image shows, the same idea as the alt attribute on a website image. Click any thumbnail to zoom in for a closer look before writing, since small chart labels or fine detail can be hard to judge at preview size. When you apply the tags, each description is embedded directly into the PDF's accessibility structure tree as a tagged Figure element linked to that specific image, so assistive technology can find and announce it — without touching or re-rendering any of the page content itself.
This is manual, human-written tagging, not an AI image captioner guessing at content. You know your document — a screenshot of a dashboard, a company logo, a diagram in a report — better than any automated system, so you write the description. On PDFs with a lot of images, the "Untagged only" filter keeps you focused on what's left, descriptions carry over with one tap when the same icon repeats, and progress is never lost to a scroll — everything runs locally in your browser, and the file is never uploaded anywhere.
Common use cases
- Making scanned reports, brochures, or slide exports readable by screen reader users
- Meeting Section 508 or WCAG 2.2 accessibility requirements for government, education, or public-sector PDFs
- Describing charts, graphs, and infographics so their data isn't invisible to non-sighted readers
- Adding descriptions to product screenshots in a PDF user manual or help guide
- Fixing accessibility gaps flagged by a PDF/UA checker like PAC or Acrobat's Accessibility Checker
One thing worth being upfront about: this covers image alt text specifically, not full PDF/UA tagging. A completely compliant tagged PDF also needs a defined reading order, tagged headings, tables, and links — a much larger undertaking that usually means rebuilding the document in dedicated authoring software. What this tool does is make the images in an existing PDF screen-reader-accessible without requiring that whole rebuild.
Related tools
PDF Alt Text Tagger — frequently asked questions
It finds every embedded image on every page of your PDF, gives you a text box to describe each one, then writes that description into the PDF's accessibility structure as a tagged Figure element linked to the image — the same /Alt entry Adobe Acrobat's "Set Alternative Text" dialog produces, just done by hand in your browser instead of desktop software.
No. Every description is written by you. The tool locates the images and gives you a place to type for each one, with a zoomable preview so you can inspect details before writing — nothing is auto-captioned or guessed.
Yes. Each description is embedded as a real StructTreeRoot Figure element with an /Alt string, linked to the image via an object reference, which mainstream screen readers like NVDA, JAWS, and VoiceOver can pick up and announce.
No, and we'd rather say that plainly than oversell it. Full PDF/UA (ISO 14289) or WCAG 2.2-level compliance means every element on the page — headings, reading order, tables, links, not just images — is tagged and linked through a complete structure tree, which normally means rebuilding the document in something like Acrobat Pro. This tool adds real, screen-reader-discoverable alt text to images specifically, without touching the rest of the PDF's structure, so it won't pass a strict PDF/UA validator like PAC on its own.
They're simply skipped — only images with a description you've typed get tagged, and nothing else in the PDF is touched. The "Untagged only" filter makes it easy to find exactly which ones still need a description before you apply.
Very rarely, a complex PDF orders its images differently than expected and one can't be confidently matched. Rather than risk attaching a description to the wrong picture, that image is skipped and flagged so you can see exactly which page it's on and try again — it never gets tagged incorrectly.
Yes — each image has a "Mark decorative" option for purely ornamental images (a divider line, a background flourish) that don't need a description. This is tracked so you can see at a glance which images you've reviewed and intentionally skipped versus which ones you haven't gotten to yet. One honest caveat: it doesn't yet write the deeper PDF/UA "Artifact" tag that guarantees a screen reader skips the image outright — that needs rewriting the page's underlying content stream, a bigger change than this tool makes today — so a decorative image is currently embedded the same way as one left blank.
Yes — while you're tagging, your descriptions and decorative markings are auto-saved locally in your browser (via IndexedDB, still never uploaded anywhere). If you come back to this tool with an unfinished PDF, you'll be offered a "Resume" option instead of starting over.
Yes. No sign-up, no daily limit, and the PDF is never uploaded — everything, from scanning for images to writing the tagged file, runs locally in your browser.