Hello everyone, I'm currently working on an AI Shopify Product Image Optimizer. It renames images for your store using the Gemini API. It also resizes, compresses, and converts them. It then uses GraphQL to replace your store images. You can pick which Product ID to start from.
It uses BASH and it's already entirely functional. Currently looking for help with an ALT text generation script using Gemini. I've been using the rename script to do so and it's obviously not good enough.
AI Shopify Product Image Optimizer - Rename using AI, Generate ALT text, Rename, Resize, Compress, Convert to WEBP.
Set your own dimension and size thresholds, gemini prompt, file text appends and prepends, and webp quality.
Batch replace over the 250 limit using last product processed cursor.
You can read the full article here: https://seo.topmiamisoftware.com/shopify-product-media-seo-ai-image-optimizer/
If anyone feels like they can contribute to the Github repo under the MIT License then please go ahead. Even if it's to open up issues, you'll be helping tremendously.
https://github.com/topmiamisoftware/shopify-seo-image-ai-optimizer
Top comments (3)
Nice tool! The ALT text generation and WebP conversion side is really well covered here. One complementary angle worth thinking about: even after you've nailed the technical optimization side, visual inconsistency across product images is still a major conversion issue. Merchants often pull photos from different suppliers and end up with mismatched backgrounds, varying aspect ratios, and inconsistent framing.
I built an app called Eye Catching (apps.shopify.com/beautiful-brands) that tackles that side — standardizing image backgrounds, adding overlays, and applying badges across the entire product catalog. Think of it as the visual consistency layer that pairs well with the SEO/technical optimization work you've described here.
Disclosure: I'm the developer. But the two approaches solve adjacent problems and complement each other well.
Some comments may only be visible to logged-in visitors. Sign in to view all comments.