A tee is sold by two things a flat-lay can't show: where the print actually sits on a chest, and how the cotton falls once a body is inside it. Upload a flat photo of your t-shirt and CatalogX generates a photorealistic on-model image with the artwork intact and the drape believable — in about 30 seconds.
Most AI image tools treat a printed slogan as decoration and redraw something that looks roughly similar. For a seller, that isn't a cosmetic flaw — it's a wrong product image.
Ask any t-shirt seller what goes wrong with generated mockups and you'll hear the same three complaints. The slogan comes back with garbled or invented lettering. The chest print drifts upward toward the collar or downward toward the hem. And a small left-chest logo balloons into a full-front graphic because the model had no idea it was meant to be small.
CatalogX handles this as a distinct step rather than hoping the render gets it right. Before generation, your uploaded tee goes through a vision pass that locates the printed area as a bounding box and performs OCR on any text inside it. The generation instructions then carry that information forward: this artwork, at this position, at this relative size, with this exact wording. A left-chest embroidery stays a left-chest embroidery. A centre-chest illustration keeps its vertical position relative to the collar and hem. Slogan text keeps its words.
That single behaviour is the difference between a mockup you can list and a mockup you have to explain to a customer who received something else. It matters most for the categories that dominate Indian t-shirt catalogues right now — anime and gaming graphics, regional-language slogans, band and film references, minimal typographic prints, and the small-logo basics that boutiques private-label.
Every garment type in CatalogX carries its own instruction set. For tees, the rules are about relaxed casual fit with natural fabric drape, sleeve length and neckline matching the reference exactly — crew, round or V — soft cotton texture with realistic creasing at the waist and underarms, a hem sitting naturally at the hip line, and prints, logos and graphics preserved as shown. Those aren't generic prettiness instructions; they're the specific failure modes of tee renders, written down.
The creasing rule deserves a mention on its own. A t-shirt rendered without creases reads as plastic — buyers can't articulate why, but they scroll past. Soft cotton bunches slightly above the waistband, gathers under the arms, and shows a small horizontal fold where the chest meets the ribcage. Getting those in the image is most of what makes a render look photographed rather than generated.
Indian marketplaces reward filled image slots. Here's a shot list that fills them, and the CatalogX settings that produce each frame.
| Slot | Camera frame | Pose & angle | What this image proves to the buyer |
|---|---|---|---|
| Main listing image | Full body | Standing, front, hands by side | True length on the torso and where the hem lands on the hip — the single biggest source of "shorter than expected" returns on tees. |
| Print hero | Waist-up | Standing, front, arms relaxed | The graphic filling enough of the frame to be legible in a mobile search thumbnail. |
| Print detail | Close-up | Product-view | Print texture and edge quality — whether it's a soft-hand DTG print or a heavier screen print sitting on top of the fabric. |
| Shoulder & sleeve | 3/4 | Standing, 45°, hand on hip | Sleeve length, shoulder seam position, and whether the cut is drop-shoulder or set-in. |
| Back view | Full body | Standing, back angle | Back prints, neck tape, and the fact that the tee has a finished back at all. |
| Lifestyle / social | Full body | Walking or leaning, urban or café background | Context for Instagram, WhatsApp catalogues and Meesho reseller shares, where a studio-white image travels badly. |
Fit is a separate control from garment type, so the same design can be rendered as four distinct silhouettes.
The streetwear default. The shoulder seam sits below the actual shoulder and the body hangs wide — a regular-fit photo on an oversized listing is a misrepresentation buyers notice immediately.
The safe default for basics and printed graphics. Enough room through the chest to read the artwork flat, without the pooling at the waist that an oversized render produces.
For gym, athleisure and fitted-cut listings where the selling point is that the tee follows the body rather than hanging off it.
Between regular and oversized. Useful for cotton and slub fabrics where the fall is soft but the cut isn't deliberately outsized.
Generate your bestselling design once as regular and once as oversized with the same model, pose and background. You now have two comparable listings, or two images inside one listing that let a buyer choose a fit without guessing. The uploaded flat-lay never changed.
T-shirt catalogues are unusual: the garment barely changes, the artwork changes constantly. That's exactly the shape of problem bulk generation is good at.
If you sell print-on-demand or drop new graphics weekly, your real cost isn't photography quality — it's photography throughput. Booking a model to shoot forty designs on the same white blank is absurd, so most sellers fall back on flat digital mockups. Those are cheap and they look it, and on a marketplace grid they sit next to competitors' on-model images and lose.
Bulk mode accepts up to fifty garment images in a single run against one chosen model, one pose and one background. Forty designs become one run, not forty sessions. Because the model face, lighting and framing are held constant, the output has the property that actually matters commercially: a storefront that looks like a catalogue rather than a folder of unrelated mockups pulled from different tools.
The library covers 21 male and 44 female models across casual, modern, athletic, mature, plus-size, teen and regional looks. For a men's graphic tee brand, picking one male model and staying with him across the season gives your catalogue a face — the same reason apparel brands rebook the same model. For unisex or women's fits, generate the same design on both a male and a female model and use both images in the listing; buyers of unisex tees consistently ask which one it's cut for.
Backgrounds range from plain studio white for marketplace main images through minimal studio colours, urban brick, street, rooftop, café and gym settings. A gym background on an athleisure tee and a café background on a slogan tee do different merchandising work from the same source file.
Generate at HD, 2K or 4K depending on where the image is going. Marketplace zoom rewards higher resolution — Amazon India only enables zoom above a size floor — while social crops rarely need it. Export presets then handle the shapes: 2000×2000 square for Amazon India and Flipkart, 1080×1440 portrait for Myntra, 1500×2000 for Ajio, 2048×2048 for Shopify, and 1080×1350 for an Instagram feed post. One master image, every channel, no manual cropping.
No studio, no model booking, no shoot day. A phone photo of the tee laid flat under even light is enough.
Flat-lay, hanger or mannequin. Lay it smooth and shoot straight down — the flatter and squarer the print reads in your source, the cleaner it renders on the model.
Choose T-Shirt as the garment type, pick regular or oversized, choose a model from the 65-strong library, then set pose, camera frame and background.
About 30 seconds per image. Run the shot list above, then export each channel's dimensions with one click.
A crisp source image makes a better render. Remove the background, flatten the creases from the shipping fold, and upscale a low-res supplier photo before you generate.
That is the part CatalogX treats as a separate problem. Before generating, a vision pass locates the print area on your uploaded tee as a bounding box and reads any text inside it, so the render is instructed to reproduce that artwork at the same position and scale rather than improvising a similar-looking graphic. Chest-centre prints stay chest-centre, left-chest logos stay small and off to the side, and slogan text keeps its wording instead of turning into invented letterforms.
Yes. Fit is a separate control from garment type, with regular, slim, oversized and relaxed options. Generate the same tee twice — once regular, once oversized — and you get two listings, or two images inside one listing, from a single flat-lay. This matters for streetwear catalogues where the drop-shoulder oversized fit is the actual product and a regular-fit photo would misrepresent it.
Plain tees are the easiest case and often the most valuable one. With nothing to look at but colour, neckline and hem, a buyer needs the model shot to judge how the cotton sits across the chest, where the hem lands on the hip, and whether the collar rib looks substantial. Generate a plain tee across three or four colourways with the same model and pose and your colour options page becomes genuinely comparable.
Four to six for most Indian marketplaces. A workable set is a full-body front on studio white for the main slot, a waist-up front that lets the print fill the frame, a close-up of the print texture, a back view, a 45-degree angle showing sleeve and shoulder seam, and one lifestyle shot on an urban or café background for social and secondary slots.
No. Bulk mode accepts up to 50 garment images in a single run against one chosen model, pose and background, so a hundred designs on the same blank body become two runs. Keeping the model and framing constant is what makes a print-on-demand storefront look like a catalogue instead of a folder of unrelated mockups.
Usually only resizing, and that is built in. Export presets output the master image at Amazon and Flipkart's 2000x2000 square, Myntra's 1080x1440 portrait, Ajio's 1500x2000, Shopify's 2048x2048 and Instagram's 1080x1350 without manual cropping. Generate once at HD, 2K or 4K, then export the sizes each channel needs.
Upload one tee, run the six-image shot list, and judge the print fidelity on your own artwork.