I will not invent seven app names to fill a list. Here are the seven things I would actually compare.
One, capture quality on real products. Shiny, transparent, and very dark objects are the hard cases. Test with your actual catalog, not the app's demo shoe. Two, lighting forgiveness. You are scanning in a stockroom, not a studio. The app should cope with that. Three, export formats. GLB and USDZ, no exceptions, or your models will not run everywhere.
Four, processing time. Some apps process on-device in minutes. Others upload to the cloud and make you wait. Across hundreds of products, that difference is a week. Five, editing tools. You will need to trim backgrounds and fix holes. A scan app with no cleanup tools is half a product. Six, cost per model at your volume. Per-scan pricing that looks cheap for ten products gets expensive for a thousand. Seven, store integration. Does the finished model flow into a 3D asset production SaaS pipeline or your product pages without a manual export dance every time.
Compare the seven criteria, pick the one that survives your actual products and your actual volume, and remember the scan is the start, not the finish. Every model needs a human review pass before it meets a customer. A mobile app to measure AR performance will later tell you whether those models earned their keep, but the review pass comes first.
Need help setting up the capture-to-page pipeline? I consult e-commerce brands on AR implementation at arcommerce.fyi.