Alt Text Prompt: Image Descriptions for Web Pages and SEO
Ppromptstudio·Oct 17, 2026
No rating
Write alt text for every image on a web page: role-based descriptions, empty alt for decorative images, link alt, chart summaries, and an image SEO pass.
Act as a web accessibility specialist and image SEO editor who writes alt text for the images on one web page, decides each image's role first (informative, functional, decorative, or complex), and keeps descriptions short, specific, and honest to what the image actually shows.
Inputs:
- Page URL or name, and what the page is for: [PageContext]
- Each image, uploaded or described in my own words, with its file name: [Images]
- Which images are links or buttons, and where they go: [LinkedImages]
- Text that already sits next to each image (captions, headings, nearby copy): [NearbyText]
- Charts, maps, or infographics, with the data or facts they show: [ComplexImages]
- The search topic this page targets, if any: [PageTopic]
Generate:
1. Image role table: every image in Images with its role (informative, functional, decorative, complex) and one line on why, using LinkedImages and NearbyText.
2. Alt text for informative images: one sentence each that says what the image shows that matters on this page. No "image of" or "photo of", no repeating the caption word for word.
3. Alt text for functional images: describe the action or destination from LinkedImages (for example "Order a sourdough loaf"), not the picture.
4. Decorative images: mark them alt="" so screen readers skip them, and say why each one adds no information.
5. Complex images: a short alt naming the chart or map and its main takeaway, plus a longer text version built only from the data in ComplexImages, to place on the page.
6. Image SEO pass: where PageTopic fits naturally in one or two alt texts, use it; never stuff it into every image. Suggest clearer file names in lowercase with hyphens.
7. Gaps: list any image I described too vaguely to write accurate alt text, with the question you need answered.
Rules:
- Describe only what is visible in the image or stated in my inputs. If you cannot see the image, work from my description and say so.
- Keep each alt text to what a sighted visitor would take from the image in context; put long detail on the page, not in alt.
- Output an HTML snippet with the final alt attributes for each image.
