AI Product Images on Shopify, ikas and Ticimax: Bulk Upload and Variant Workflow

The hard part of AI product imagery is no longer generating it; it is moving hundreds of frames into the store: which size, which format, which variant, which order? This article lines up the image rules of three widely used storefront platforms (Shopify globally, ikas and Ticimax in Turkey) and gives the step-by-step workflow that connects AI output to a bulk upload pipeline. Rules were compiled from the platforms'' own documentation; they change over time, so the current help page in your admin panel has the final word.

Three platforms at a glance

Shopify

• Recommended 2048 x 2048 pixels, square (1:1); upper limits 5000 x 5000 pixels and 20 MB.
• Zoom only works above 800 x 800 pixels.
• Up to 250 images per product; each variant (color, size) can carry its own image.
• Upload JPG/PNG; served automatically as WebP to supporting browsers.
• CSV bulk upload is URL-based; a separate "Variant Image" column maps variant images and "Image Position" sets order.

ikas

• Accepts jpeg, jpg, png and webp; 10 MB ceiling per file, under 4 MB recommended in practice.
• Up to 10 images and/or videos per variant on variant products.
• Excel/CSV bulk upload: the image URL column takes addresses ending directly in an image extension (.jpg, .png, .webp); multiple images separated by semicolons (;).
• For variant images, create the variants first, then attach images per variant.

Ticimax

• Bulk image upload runs through Excel: Products > Product Management > Import > "Bulk Image Upload with Excel".
• Matching by one of four keys: product card ID (parent products), product ID, stock code or barcode (variants); the image address goes in the RESIMURL column, optional ordering in a sequence column.
• "Upload images to variations" and "Delete previous images" are options at upload time; XML integration also carries images in bulk.
• No official size or format ceiling in the documentation; your theme''s product card ratio (3:4 portrait on most Turkish themes) and files under 1 MB are the safe standard.

The common denominator: URL-based upload

On all three platforms, bulk image upload has the same backbone: images sit at a web address (your server, a cloud bucket or an image CDN), each spreadsheet row carries that address, and the platform pulls the image and copies it into its own storage. So the first decision of an AI workflow is "where will the images live"; downloading output to a laptop and uploading product by product collapses at hundreds of SKUs.

From AI output to the store: step by step

• 1. Set a standard. 2048 x 2048 square for Shopify; around 1200 x 1600 portrait (3:4) for most ikas and Ticimax themes. Long edge 1600 to 2048 pixels, 200 to 400 KB webp: the right band for both quality and speed.
• 2. Set a naming rule. File names should carry stock code and variant: SKU1234-black-1.webp, SKU1234-black-2.webp. The spreadsheet can then be derived from the names, and no image lands on the wrong variant.
• 3. Generate the AI images. On-model frames and color variants come from the plain frame via an AI virtual model; each color variant gets its own on-model frame, uploaded as the variant image. Verify each variant''s color against the fabric swatch.
• 4. Host the images. A cloud bucket or CDN; every file needs a direct address ending in the extension (ikas and Ticimax require it).
• 5. Build the spreadsheet. Shopify: Handle + Image Src + Image Position + Variant Image; ikas: the image URL column (semicolon-separated); Ticimax: stock code + RESIMURL + sequence. The first image is always the plain main frame, followed by on-model and detail frames (the marketplace layout logic applies here too).
• 6. Test with a small batch. Upload 10 products, check variant mapping and order, then run the full list.

Variant images: the color the customer sees is the color they picked

On all three platforms the variant image swaps the main image when a customer picks a color; the conversion effect is large, because the customer immediately sees "how does this color look on me". In an AI workflow that means a separate on-model frame per color, which is exactly what deriving every color''s model image from one sample is for. Color accuracy is as critical in retail as in wholesale; the returns connection is in the return-rate guide.

Speed: heavy images slow the store

Shopify converts uploaded JPG/PNG to WebP automatically; ikas accepts WebP uploads directly; on Ticimax you control file size yourself. In every case, compressing images to the 200 to 400 KB band at generation time protects page speed; slow pages hurt both conversion and Google ranking. Use a separate close-up frame for detail instead of uploading a giant main image.

Common mistakes

• Image addresses that do not end in the extension (share links): ikas and Ticimax cannot fetch them.
• Square images on a 3:4 theme or portrait images on a square theme: cropped products, empty margins.
• Uploading without variant mapping: every color shows the same image and customers add the wrong color to cart.
• Ticking "delete previous images" carelessly: the old catalog disappears.
• Skipping the post-upload check: listings with scrambled order go unnoticed for days.

Platform rules look technical, but the core is simple: one standard, one naming rule, a URL-based store and the right spreadsheet template. With those four in place, moving hundreds of AI images into a store is an afternoon''s work. For the cost side of imagery, see the pricing guide.