Fill in product data from images. Review it inside Shopify.

1tm AI Product Info is the Shopify app from 1tm solutions GmbH for AI-assisted product data work. It drafts titles, descriptions and features from the product data and images you already have. Merchants review the suggestions and apply the values they want, inside Shopify admin. The app is publicly available in the Shopify App Store.

Both run the same analysis engine. The playground is where you test your own images and attribute templates. The Shopify app reads existing products together with their category and variant context, and adds the review and apply step in admin. The Shopify app needs no separate API access and no additional web app subscription; you manage its plan through Shopify.

You select one product, or up to 50 for a batch analysis. Google Gemini processes the existing product data and images; matching web sources can add to that. You then compare existing and proposed values and apply your selection per product. Runtime depends on scope and research. A queued analysis is not yet an updated product.

On the product page through “More actions” and “Enhance with AI”, in Shopify's product list through “AI Enhance Selected”, or on the app's own Products page through “Analyze” or “Analyze N products” for your selection. An additional AI enrichment block in admin is optional. You do not need your own Google API key.

Yes. Existing copy and product data can be enough, for downloads or handmade products for instance. A missing barcode is not an error. Clear images help with visible properties; ingredients, material composition, compatibility or certifications need reliable stated information. The app can only work from the information available.

The analysis takes the product type, any existing category information and the variants into account. That covers fashion, spare parts, food, furniture or digital offerings, among others. Additional fitting features can be suggested. Complete recognition of every property is not promised; where evidence is missing or contradictory, fields can stay empty.

Up to 50 variants and five images are considered per product. With a larger variant count a notice appears. Shared features should hold for every variant; individual colours, sizes and barcodes must not be generalised. Variant SKU, options and barcodes are not changed. With several variants, or variants that were only partly read, there is no global GTIN suggestion.

By default, existing product fields, translations and feature values are kept. Only “Overwrite existing values” permits the selected values to be replaced. Tags are added to. Read the apply notices: a protected value can stay unchanged even though you selected the suggestion.

Only if you enable auto-apply in settings. Qualified values at or above your threshold, 85 % by default, can then be written automatically when finished results are fetched, including for batch analyses. Existing content stays protected; tags can be added to. Without a confidence score nothing is applied automatically. For descriptions, only the first qualified language version is currently processed automatically.

Confidence is the model's estimate, not a measured hit rate. “Not assessed” marks a missing score; those suggestions are not preselected. Sources appear only where they are available and attributable. A product-wide web match does not evidence every field. Check anything that matters against your own product documentation.

Title, description, vendor, product type and tags, plus further description languages. Selected features are stored as linked metaobjects, and available provenance as metadata. The app does not write a Shopify taxonomy category, a GS1 GPC assignment, prices, inventory, images or the separate SEO fields. A GTIN suggestion stays a metafield and changes no variant barcode.

Additional features are linked to the product. A visible feature table needs theme support; filters additionally need matching fields and configuration in Shopify Search & Discovery. The optional AI enrichment block is an admin block. Native descriptions, by contrast, appear in your theme's existing text areas.

Free covers 10 completed analyses per UTC calendar month, with no base price and no additional usage charges. Standard costs USD 19 per 30 days, includes 100 completed analyses per UTC calendar month, and costs USD 0.10 per further analysis. An app usage cap, USD 50 by default, applies on Standard only. The plan confirmed in Shopify is the binding one. Under “Billing” you can move to Free through “Change or cancel plan” and end recurring charges. Shopify shows when the change takes effect before you confirm.

A new successful analysis run counts again, including for the same product and including when you apply nothing afterwards. Reopening the same result and applying it do not count again. More description languages stay one run. Failed runs and explicitly empty results do not count as a completed analysis. The app's usage display follows the UTC calendar month, which can differ from your Shopify billing period.

Descriptions can be produced in German, English, French, Spanish, Italian, Dutch and Portuguese. Select the languages you want that are published in your shop. Titles, tags and features each have one visible language version. If your primary shop language is unsupported, or a fetch fails, English can appear as a fallback. The app interface is English.

You can choose Professional, Casual, Technical or Marketing, and highlight one evidenced particularity under “Product Highlight”. The app can take existing category information and additional fitting features into account. An editor for freely defined attribute templates is not currently part of the merchant interface.

The publication status is not changed. On a product that is already published, applied copy can become visible immediately. Test on a draft product of your own first. There is no undo function of our own; back up content that matters before you deliberately overwrite it. Uninstalling does not revert stored changes.

For the analysis, product images and the necessary product information go to Google Gemini. The workflow needs no order or customer lists. Your shop connection, settings and usage data are processed to run the service. Use material you are permitted to use. Processing, retention and deletion are set out in the privacy policy, which is linked from the guide.

The German and English guides cover setup, review, applying and the common errors. If a question is still open, write to contact@1tm.solutions with your shop domain, a product link, the time it happened and the exact error message. After a save error, check the product's current state first, since partial changes may already be stored. Please do not send credentials.