Add accessible alt text to social media images fast: where the field lives on each platform, how to write a strong description, and what to skip.
Last verified · 2026-11-05 · by Moe Ameen
Alt text is the written description a screen reader speaks aloud in place of an image, and on social media it is a one-sentence field that most creators leave blank or let the platform auto-fill badly. Filling it well is quick once you know where each platform hides the field and what a good description actually sounds like — it serves the roughly 1.3 billion people the WHO estimates live with a significant disability, many of whom navigate feeds by screen reader, and it quietly gives the platform's search and recommendation systems one more plain-language signal about what your image shows.
This is the hands-on version, in order: decide which images need a description, write the description before you touch the field, add it in each platform's native alt-text control, handle the platforms that auto-generate it or don't support it, transcribe any words baked into the image, and review anything an AI drafted for you before it ships. The limits and field locations shift between apps, so the workflow matters more than memorizing a number — but the bar for a good description is the same everywhere.
Alt text is a core requirement of the Web Content Accessibility Guidelines (WCAG) and can be part of legal obligations under laws such as the ADA or Section 508, depending on your organization, industry, and jurisdiction. Some public bodies set hard deadlines — e.g. a City of Los Angeles guidance page requires new social content to meet WCAG 2.1 Level A and AA from April 24, 2026. Adding alt text alone does not guarantee compliance; pair it with captions, transcripts, strong color contrast, and clear link text, and check your own organization's policy.
Alt text is quick to write when you already know exactly what the image shows — and slow when you're reverse-engineering a stock photo someone else shot. That's the quiet advantage of generating your images instead of sourcing them, and it's where [Kompozy](/) fits this workflow. Kompozy is a full AI content generation and multi-platform publishing engine, not a repurposer: you generate the image set from a known topic and [Persona Brief](/glossary/persona-brief) — Photo Posts, face-locked Persona Photos, Carousel Posts and infographics built pixel-exact through [HyperFrames](/glossary/hyperframes), and Quote Graphics — so when you reach the alt-text field you already have the subject, the on-image text, and the point of the shot in front of you. Writing an accurate description becomes a ten-second confirmation, not a research task.
The place that confirmation happens is the per-post review pipeline. Kompozy routes every generated piece through an approve-before-publish gate, and that gate is exactly the right station to add each image's alt text — the same moment you're already checking the copy and the brand look, image by image, before anything ships. Quote Graphics help here specifically: their text is rendered as a real server-side layer, so the words you transcribe into the alt field match the card exactly rather than being reverse-read off an image model's output. Then [Autopilot](/glossary/autopilot) schedules and fans the approved set across the eight social platforms plus blog and email, so the review-gate pass covers your whole distribution instead of one upload.
Be exact on the boundary this guide draws: alt text is per-image and per-platform, and the native field in each app — Instagram's "Write alt text," X's ALT badge, LinkedIn's alt-text control — is still where the description is entered and is still a human judgment about what matters in the shot. Kompozy removes the slow part (producing a legible, on-brand, known-subject image set and giving you one review surface to describe it all), not the human part (writing a true, specific description per platform). Starter runs $199/mo (5,500 credits) for a solo creator shipping a steady image set; Pro is $499/mo (18,000 credits) for a team producing across every surface; Enterprise is available through Contact Sales.
Aim for about a sentence or two — roughly 125 characters is a useful working target — and only go longer when the image genuinely needs it, such as a dense infographic. Most platforms' native fields allow far more (X, for example, allows up to 1,000 characters per image), but a concise, specific description read aloud beats a long one. Put the most important detail first so it survives any truncation.
No. The caption is public copy everyone sees — commentary, hashtags, links, a call to action. Alt text is a description of what the image actually shows, read by screen readers and used as a plain-language signal by platform search. Keep them separate: don't paste your caption into the alt field, and don't hide a call to action in alt text. Describe the image in the alt field; sell the post in the caption.
Add it wherever there's a field: Instagram, Facebook, X, LinkedIn, Pinterest, Threads, Bluesky, and TikTok's photo-mode posts all support native alt text. Instagram Stories, regular TikTok videos, and YouTube have no per-image alt field, so there you move the information into on-screen text, captions, or the description instead. Because alt text is written per image, the same photo posted to several platforms needs the description entered in each app — it does not carry over.
Accessibility is the primary reason and the one that matters most. But alt text is also a free, mostly-blank surface that platform search and recommendation systems read as an extra plain-language signal about what an image shows, and it helps people on slow connections or with images turned off. The catch is that it only helps when it honestly describes the image — keyword-stuffed or inaccurate alt text hurts both the user and your credibility.
Facebook auto-generates alt text with AI for images that don't have a human-written description, which is why you'll see strings like "May be an image of one person and text." Those guesses are frequently inaccurate or vague. Override them: click the image, choose Edit, then "Alternative text," and replace the machine guess with a description of what the image actually shows.