> ## Documentation Index > Fetch the complete documentation index at: https://docs.trychannel3.com/llms.txt > Use this file to discover all available pages before exploring further. # Images High-quality images are returned with each product. We classify every image with AI and attach metadata on the `ProductImage` type — including `shot_type`, generated `alt_text`, and an `is_cleaned_image` flag. ## Shot types Every image is classified into a `shot_type` so you can curate what shoppers see. Building a scrollable grid? Prefer `hero` images. Want a mood-board feel? Try `lifestyle`. Filter out `size_chart` from carousels. Use `on_model` for fit and virtual try-on flows. Each product can have multiple images across types. Read `shot_type` on any image in the `images` array: ```typescript TypeScript theme={null} const heroImages = product.images.filter((img) => img.shot_type === "hero"); const withoutCharts = product.images.filter( (img) => img.shot_type !== "size_chart", ); ``` Examples below are from major retailers in our catalog — `RelevanceLevel.MAJOR` (relevance score 4), with AI-analyzed, displayable images.

Shot type	Examples
`hero` Primary product shot — usually front-facing on a clean background.
`lifestyle` Styled or environmental context — editorial settings and mood shots.
`on\_model` Model wearing the product — useful for fit and virtual try-on.
`detail` Close-up of fabric, stitching, hardware, or texture.
`angle\_view` Side, back, or alternate viewing angles.
`flat\_lay` Product photographed flat from above.
`in\_use` Product actively being used, worn, or opened in context.
`scale\_reference` Size context — dimensions annotated or shown next to reference objects.
`packaging` Product in retail box, bag, or packaging.
`size\_chart` Sizing tables or measurement diagrams.
`product\_information` Specs, features, or infographics about the product.
`merchant\_information` Retailer marketing — banners, warranties, and store messaging (not the product itself).

Because images come in various styles and aspect ratios, we also provide **cleaned images** — square aspect ratios and uniform backgrounds that make product grids look consistent. ## Cleaned images Cleaned images have: 1. Square aspect ratio 2. Uniform background 3. Centered product They're indicated on the API with `is_cleaned_image: true` on the `ProductImage` type. Take [nordstrom.com](http://nordstrom.com) on our dashboard. Before cleaned images, the product grid is good, but bad aspect ratios stick out and on-model shots break up the layout. A Nordstrom product grid before cleaned images, with mismatched aspect ratios and on-model shots

A Nordstrom product grid before cleaned images, with mismatched aspect ratios and on-model shots

With cleaned images, every tile is a square lay-flat shot — on-model photos are replaced, and narrow crops (like sunglasses) are padded to square. The same Nordstrom product grid after cleaned images, with uniform square tiles

The same Nordstrom product grid after cleaned images, with uniform square tiles

## Prefer cleaned images in grids When displaying search or similar-product results, pick the cleaned image when one is available: ```typescript TypeScript theme={null} const heroImage = product.images.find((img) => img.is_cleaned_image) ?? product.images[0]; ``` This works on any product response that includes `images` — search, lookup, and product detail all expose the same `ProductImage` model. Visual search is a common place to surface cleaned images in a grid. Combine color filters with cleaned images for palette-matched, uniform grids. Channel3 components can prefer cleaned images in product cards automatically.