10mShop

Images and videos

The asset picker, free stock search, AI generation — and the label every AI picture carries.

Esta página aún no está traducida. Se muestra en inglés; todo lo que describe funciona igual en tu idioma.

Every image slot in the designer offers the same four sources: upload, your assets, free stock search, or AI generation. Image slots can also hold videos. An upload takes a JPEG, PNG, GIF or WebP picture, or, with the Pro plan, an MP4, WebM or MOV video of up to 50 MB; all of a shop's own videos together hold up to 1 GB. Stock videos from the search are free on every plan.

The asset picker

Called Design Assets. Two groups of tabs:

Library — All and three tabs, pictures and videos alike:

  • Uploads: your uploads and the pictures and films the AI made for the design.
  • Web: the stock pictures and videos your design uses.
  • Products: every picture and video of your products, yours and the AI's.

A small choice beside the count narrows every library tab to Images only or Videos only; All shows both. A tab shows 30 files at a time: Previous and Next below the library turn the page.

Search — Find photos and Find videos, both free Pexels stock search. Your own videos are in the library tabs.

Each asset shows a source badge — Unsplash, Pexels, AI, Product, Web — and, if it is used in the current live or preview version, a Live or Preview chip. Hovering shows which block and page it is used on. AI marks every picture and film the AI made, product pictures included.

Managing your assets

Theme & Assets in the designer toolbar opens the same library in a wider window, without an image slot to fill: its Assets tab, beside Theme.

  • Upload takes several pictures or clips at once; for a clip the designer first asks what it shows, and whether it was made with AI.
  • Create with AI makes a new picture or a film; either joins your uploads (Films from the AI).
  • Add to my uploads under a find of Find photos or Find videos keeps a copy in your shop, under Uploads; a find you already keep says In your uploads. A stock video takes the name of its Pexels page as its sentence, and keeping one needs the Pro plan, like uploading a video.
  • A click on a picture opens what the AI sees on it, with your own sentence.
  • The bin on an upload deletes it.

The AI places stock videos by itself where they fit — hero backgrounds, mostly — when it generates a storefront. For a new block or page it looks at the preview of a few videos found and takes the one that shows best what belongs there.

What the AI sees on your pictures

Under every upload, AI picture and product photo of the library the picker shows one sentence: what the AI sees on it, in your shop's language. The AI looks at a new picture once, the first time the picker or the picture choice of a new block or page is open; until then the line says it is looking.

Click the sentence to add your own: "Forest-green earrings, our favourite gift". The AI reads it along with its own whenever it picks your pictures for a new block or page, so your words steer which picture goes where. A clip has only your sentence — the AI cannot watch videos —, so the designer asks for it when you upload a clip, and the clip only goes up with one. With it, the AI puts your clip where a new block wants a video — a hero's background, a video block — when it shows what belongs there. An older clip without a sentence is marked, and the AI never takes it. A clip in a product's gallery needs no sentence: the AI knows it by its product's name and offers it with the product photos; your sentence under it tells the AI more.

The search above the library finds your pictures by these words, yours and the AI's, in your language and in English: "earrings" finds the earrings, product photos included.

Your shop reads these sentences out, too. Where a block has nothing but its headline or a title to say about a picture, screen readers for blind visitors — and search engines — get the sentence of what is on it instead, and a clip's sentence for a clip. Words a block has for the picture itself, such as the alt text of a gallery picture, stay, and a logo, a portrait or a product keeps its name. A picture the AI has not looked at yet, and every stock photo, keeps the block's words.

The gallery hero and the two standalone gallery blocks (masonry columns, horizontal strip) have a grid editor in the Images tab: add, replace, remove, reorder or AI-generate up to the block's maximum.

The standalone gallery blocks take up to 16 images regardless of how many the AI originally created, and their grid editor also edits each image's alt text and caption.

  • Captions appear in the shop while the block's Show Captions toggle is on.
  • An image with no caption falls back to its alt text, so the toggle always shows something.

The horizontal strip gallery has a Header Alignment option (left / center / right) that places the headline and — on desktop, when the images fit in one row — the strip itself. On phones and tablets the strip always starts at the left gutter and scrolls sideways, resting on the same side padding as the text above it.

Your own products in an AI picture

The AI only knows your products when you show them. Under the description of an AI picture, Put my products in the picture lets you pick up to four of your products; their main pictures go to the AI as a model, so a living room "with three of my paintings" shows your paintings and not invented ones. This works in every image slot, in the gallery grid and in the small editor on the page itself. Only products with a picture are listed.

The AI paints the products after their pictures: very close, but not pixel for pixel, so small differences are possible. In a Picture with products block the picked products are listed first when you set the dots; the AI never sets the dots itself.

AI-generated pictures carry a label

Every image the AI creates — an image slot in the designer, an AI product picture, a product's AI sales picture — shows a small "AI-generated" label on the storefront, in your shop's language.

It is always on and cannot be removed: the EU AI Act (Article 50) requires the disclosure. The files also carry a machine-readable "made by AI" mark in their metadata (IPTC, the field Google and Meta read) and OpenAI's invisible SynthID watermark; AI films carry the same mark.

The one exception is the AI-created logo, which carries no label.

Films the AI makes carry the same label. A clip you upload carries it when you tick Made with AI — when you upload it, or later under its sentence in the library. Nothing tells reliably whether a clip was made with AI, so your tick decides; when the file says so itself, as some AI tools write into it, the box is already ticked.

What AI generation costs

Image generation consumes AI credits, like every other AI feature. The balance at the bottom of the sidebar updates by itself after every AI action.

En esta página

Images and videos — 10mShop