Modulify

Images

Get real photography, generated artwork or your own pictures into a site.

On this page

You do not manage images in a library. You ask for them in chat, and Modulify puts the files into the project and wires them into the page in the same turn.

There are three routes, and they cost different amounts.

Stock photography, free

Searching real stock photography costs nothing, and it is the first choice for anything that exists in the real world: people, teams, offices, products, food, interiors, places, nature, events.

Add photos to the about section. A small team working in a bright open-plan office.

Modulify searches with the query it builds from your request, downloads what it finds into the project, and references the local path. A whole page of images is fetched in one go, up to 10 searches per request, each with its own subject.

Say what the shot should be, not just the topic. "Barista pouring latte art" returns something usable. "Coffee" returns filler.

You can steer the kind of imagery: photographs are the default, and you can ask for illustration or 3D artwork instead. Still images only, never video or animated files.

AI generated images, 1 credit each

When no real photograph would do, Modulify generates one. This is for bespoke illustration, stylised or surreal scenes, brand-specific visuals, abstract art, icon sets, diagrams and conceptual artwork.

Generating costs 1 credit per image, charged to the workspace. Because it spends money, Modulify asks before generating anything you did not clearly request, and reads a vague "add some images" as stock photography instead.

Describe the picture the way you would brief a designer. Subject and what it is doing, the setting, the framing, the lighting, the palette in concrete colours, the mood, the medium, and how much detail. Say what you do not want when it matters.

Generate a hero illustration: a delivery drone over a city skyline at dusk,
wide shot with the drone low and left, warm orange into deep indigo,
flat vector, generous empty space across the top for a headline.

Words inside the picture

The image model renders written text accurately, which matters for posters, badges, packaging and app mockups. Quote the exact string you want and say where it sits. Keep it short, a few words rather than a paragraph. If you want no lettering at all, say so.

Shape and resolution

Aspect ratio is one of 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9 or 21:9. The default is 16:9.

Resolution is either 1K or 2K, and 1K is the default. The size changes pixel dimensions only, never the composition and never the price, so both cost the same 1 credit. 2K is for a full-width hero that has to stay crisp on a large screen. There is nothing above 2K.

Up to 10 images per request, and the whole page is generated in one batch so the images read as a set.

AI edited images, 1 credit each

Editing takes a picture that already exists and changes it, keeping the original subject, composition and identity rather than inventing a new picture. It also costs 1 credit per image.

Use it to change the season, time of day, weather or mood, recolour or restyle something to match your brand, remove or add an object, clean up or replace a background, extend or recompose the frame, or make a set of images look consistent.

The fastest way in is the preview. Turn on the select tool, click the image on the page, then say what to change.

Make this a night shot with warm street lighting.
Keep the product, its label and the layout exactly as they are.

Naming what must not change is the part people skip, and it is what preserves the original. The vaguer the instruction, the more the model reinvents.

The source can be a picture already in the project or a public image URL. It must be a PNG, JPEG or WEBP, and a local source larger than 12MB is refused.

Editing in place is destructive

By default an edit overwrites the source file and the original is not recoverable. If you might want the original back, ask for the result to be saved as a new image rather than replacing the old one.

Your own pictures

Attach an image to a chat message and say you want it in the site. Modulify downloads it into the project and uses it.

The distinction matters: a screenshot you paste to show a problem is reference material and is not saved into the site, while your logo or a product photo is. Say which one you mean. See Attachments and context.

Where the files land

Images go into the project under public/images/, and the page references them by their public path, so public/images/hero.png is used as /images/hero.png.

Every image is a flat raster file ending in .png, .jpg or .webp. You can browse them in the File Explorer tab, behind More in the tab strip.

What none of this can do

The image tools produce one still, flat picture. They cannot make:

  • video or animation, including animated GIFs
  • an SVG or any vector file
  • an .ico favicon
  • a dependable transparent cutout

Asking for one of these gets you a plain statement of the limit and the real alternative in the same reply, not a silent substitution. For a cutout, the usual answer is a flat background colour matching the section, finished in CSS, or attaching a proper cut-out image yourself, which keeps its transparency.

Editing also does not sharpen or upscale. Rendering a small blurry source at 2K buys nothing.

When credits run short

If a batch would cost more than the workspace has left, none of it runs and the reply tells you the count, the total cost and what you have remaining. Ask for fewer images or top up.

Individual images inside a batch succeed or fail independently, so a partial failure is reported rather than silently re-run, which would charge twice.

Modulify cannot see the finished picture. It will tell you what it asked for and leave you to look, so a reply that does not describe the result is working as intended.

Next