Gather

Gather searches the image itself, not just its text. When a reference enters the library, its style, mood, palette, composition, subject, and interface elements become searchable context. So what does that actually change when you sit down to find something you saved months ago? It gives retrieval more to work with than filenames and OCR alone. Queries can be approximate, since the point is to describe the thing the way you'd describe it to a collaborator.

Collecting inspiration shouldn't turn into filing. Gather automatically describes and tags saved references so you build a searchable library without maintaining folders by hand. When the system guesses wrong, you can edit descriptions and tags and rerun analysis later. Those corrections stay attached to the reference instead of vanishing behind a recommendation feed.

Copy for AI packages a visual reference and its context into a consistent markdown handoff, including the image URL and instructions to treat it as visual direction. You paste it into an AI tool that can fetch or inspect images, such as Claude, Cursor, or v0. The same handoff is generated by the web app and the API, so the wording doesn't quietly change between surfaces. A concrete reference, in other words, usually beats another paragraph of adjectives.

Gather's browser capture lets you save a reference while you're looking at it, from a social feed, a website, or a product page. That keeps collection close to the moment of discovery, which is when most inspiration is actually found. The reference lands in the same searchable library as everything else.

The product exposes an API alongside the web app, and the handoff for AI tools works through both. That matters if you want references flowing into a bigger workflow rather than living in one browser tab. For most everyday users the web app is enough, but an API opens the door to automating the collection, tagging, and retrieval of references across your own tools without copying and pasting by hand.