Opifix

Guides

AI mockups for custom products: what to trust and what to check

An AI mockup is reliable for agreeing the look of a custom product and for pitching it. It is not reliable for colour, scale or manufacturability until a person and a factory have checked it.

Updated 3 November 2026

You can trust an AI mockup to show the look of a custom product: the shape, the composition, where the logo sits and the general mood. You cannot trust it on exact colour, real size or whether a factory can make the product until each point has been checked against a spec, a factory's opinion and a physical sample.

Most mockup problems start the same way. The picture that won the internal meeting quietly becomes the production reference, and nobody decides that it should. This guide shows where to stop that from happening.

What an AI mockup is good for

A mockup makes a decision visible. Instead of describing a product in an email, you show it, and everyone reacts to the same image.

It does these jobs well:

In all of these the question is "do we like it?", and a mockup answers that well.

Where a mockup misleads

Colour accuracy

A screen cannot show a Pantone colour. Every display renders colour differently, and the same ink looks different on cotton, plastic, metal or coated paper. A deep blue on the mockup tells you what you intend, not what you will get.

If the colour matters, for example a brand colour on a visible area, treat the screen version as approximate. Ask for a physical colour reference before you approve anything.

Scale

An image has no real dimensions. A logo can look generous on the mockup and small on the finished product, or the reverse. Proportions drift too: handles, straps and lids are often drawn a little longer or shorter than a real build would allow.

Write the measurements next to every mockup, print area included, before anyone relies on how it looks.

Whether the product can be made

This is the biggest risk. AI image tools draw what looks plausible, not what can be built. A seam may fall where no two panels could join. A closure may appear that would need a mould nobody has planned for. A material may be shown holding a sharp fold or taking an all-over print when it cannot do either.

A good-looking mockup of a product you cannot produce at your quantity or deadline is worse than no mockup. By then the team has already got attached to it.

How to check a mockup before it becomes a production reference

Do these in order. Each step removes one kind of surprise.

  1. Label the file. Until it is verified, call it a look reference, not a spec, and put that in the filename.
  2. Write down the colours. List the Pantone codes you actually want, not what shows on screen, and plan to see a physical swatch or sample.
  3. Add dimensions. Overall size, print area, logo size and any part that moves or folds, in a table beside the image.
  4. Name the materials and the print method. A mockup rarely shows whether a logo is printed, embroidered or moulded, and the method decides what is possible.
  5. Ask what has to be built. Does it need a new mould? Do the shape or material differ from a standard base? The answers change timing and minimum quantities.
  6. Get a factory's view. Someone who makes this type of product should confirm the construction before you commit to the image.
  7. Approve a physical sample. For custom products the sample is approved before mass production, and from then on the sample is the reference, not the mockup.

Keep a version history as well. When several mockups are in circulation, it is easy to send the wrong one to the factory.

How Opifix uses mockups

At Opifix the mockup is where the checking begins. AI Studio puts your logo and Pantone colours on a product mockup and says whether the product can start from a ready-made base or needs custom production, with the reason. The reasons include an exact Pantone, a colour that is not stocked, an all-over print, a print area larger than the method allows, a change of material or shape, and a quantity below the minimum.

When a product has to be created from a photo, a sketch or a description, the AI reads the construction: materials, closures, panels and whether a mould is needed. It then prepares a product sheet with four views and a first estimate that already includes margin, freight and duties. That estimate is refined when factories quote.

Factories are found by an AI research agent that only proposes manufacturers from pages it has actually checked, and the Opifix team reviews the candidates. People at Opifix check every quote, sample and order. AI drafts, people decide.

Agencies can use the same material to present concepts to their own clients. There is more on the agencies page.

How to start

You do not need a finished design. A reference photo, a sketch or a short brief with audience, occasion, brand, quantity and deadline is enough.

Tell us what the product is for and which details cannot change, such as a brand colour or a size. We will show you a mockup, say plainly what it can and cannot tell you, and list what needs to be verified before production. You can begin from the Opifix home page or go straight to the AI chat.

Pages mentioned

Drafted with AI assistance. Reviewed and approved by Marco Legnini on 3 November 2026.