Generate alt text for images on your Mac. Locally.
Create accurate, consistent alt text for website images — improve accessibility and SEO with on-device AI, without uploading a single file.
Requires Apple Silicon Mac with macOS 26
Alt text is essential — and painful to write at scale
Accessibility standards and best practices expect descriptive alt text for meaningful images, and consistent alt text supports discoverability in many workflows. But writing it by hand is slow, subjective, and hard to keep consistent as your library grows. Cloud generators can help — but they often mean uploads, variable costs, and privacy/compliance concerns.
Generate alt text privately with on-device AI
VisionTagger runs vision models locally on your Mac to describe images automatically. Process hundreds of images in one run, review and edit as you go, and export structured alt text for your pipeline — no uploads and no subscriptions required.
Bulk alt text from folders or Photos Library
Drop a folder of website images into VisionTagger — or select images from your Photos Library. Generate alt text for the entire batch in a single run.
Consistent results with presets and prompt templates
Start with built-in styles (brief, standard, detailed) or define your own rules once — length, tone, what to include or avoid. Apply the same style across hundreds or thousands of images for consistent output.
Review and refine while generation runs
Alt text appears as each image completes, so you can spot-check, edit, and iterate without waiting for the full batch.
Export a JSON mapping your pipeline can use
Export a single JSON file that maps each image path to its generated alt text — ready to wire into your build step, CMS import, or component library.
Examples
System requirements
-
macOS Tahoe 26.0 or later
-
Apple Silicon required (M1 or later)
-
For optimal performance with larger models 16GB RAM or more is recommended
-
Model storage: plan for ~4–8 GB per model (downloaded locally)
One-Time Purchase
VAT included (except US & CA)
Secure payment via FastSpring