
A 3D viewer can frame a model attractively without proving its real-world size. Shopify's modeling guidance requires correct units and actual product dimensions, and identifies incorrect AR scale as a rejection issue. Verify the delivered model beside a measured reference or the real product in the intended AR workflow.
Key Takeaways
- Apparent size in a framed 3D preview is not physical-scale evidence.
- Check model dimensions, units and the exported file together.
- Review scale in the destination device without substituting visual preference for measurement.
Start from verified dimensions
The official merchant modeling guide says the model should be built to real-world scale with correct units. Gather the real product's height, width, depth and relevant component dimensions before altering the model.
A generated lifestyle image cannot supply those measurements. The measured-dimension diagram guide explains why perspective and invented surroundings are unreliable rulers. Use actual measurements or a trusted product record.
Record the units beside each value. A bare number such as 25 leaves open whether the source means millimeters, centimeters or something else. Do not repair a dramatic mismatch by shrinking the model until it looks plausible in one screenshot.
Trace dimensions through the model and export
Have the responsible modeler inspect the working scene's units, actual bounds and relevant transforms. Then inspect the exported asset rather than assuming the working application's display proves the export preserved scale.
Keep the previous model and source project before changing them. An edit to global size can conceal a component-proportion defect. If a handle is too large relative to the body, making the whole object smaller does not fix that relationship.
For a proposed lamp model, measure the base, total height and shade separately. If its total height is correct but the shade is disproportionate, investigate geometry rather than a simple unit multiplier. This is a proposed diagnosis, not an observed model defect.
Review physical size on the receiving route
Shopify recommends that the merchant review the model in a suitable format and on an appropriate device. Its guide says the AR model placed beside the physical product should be the same size.
Use this proposed acceptance record:
| Evidence | What it establishes |
|---|---|
| Trusted measurements | Intended physical dimensions |
| Working model inspection | Scene-level units and proportions |
| Exported-file inspection | Properties of the actual delivery asset |
| Destination AR comparison | Observed size in the intended viewing route |
Keep any viewer resizing or placement behavior visible in the record. A manually adjusted presentation should not be reported as proof that the original file arrived at the correct scale.
Check appearance after scale is correct
Review silhouette, materials and labels against authorized references. A correctly sized but visually different product still fails its commercial purpose. The hand-model scale guide reinforces why a familiar surrounding object does not replace measured evidence.
Retain the accepted model, reference values, format and review result with the approved-version handoff record. Accept the asset only when its measurable dimensions and destination behavior agree with the real product, while recording any unverified device or format separately.
Related tasks
- Can an AI Lifestyle Image Prove a Product’s Dimensions?
- How Do I Use a Hand in a Product Photo Without Misleading Scale?
- How Do I Keep the Approved AI Image or Video from Getting Lost Among Revisions?
Sources and scope
- Shopify: Creating 3D models for merchants: Models should use correct units and real-world size. Shopify lists incorrect scale, where a model and physical product differ in AR, as a rejection issue and recommends reviewing the actual delivered format on a suitable device.
Based on the complete official modeling guide captured October 29, 2026. No model, unit conversion, export, device or AR placement was tested. The physical comparison procedure is proposed. No automatic model export or dimension verification capability is attributed to Imagild.
Prepared by Imagild Editorial, the company editorial team responsible for these guides. AI assisted research and drafting. This is documentation-based guidance with proposed examples, not a claim of hands-on testing or an independent product review. Report a correction.
Cover: AI-generated conceptual editorial illustration; not a real application screenshot, observed output or before-and-after test.