AI shopping assistant

A fashion assistant built around the image you saved.

Remode starts from the actual look: a photo, screenshot, TikTok frame, Pinterest pin, or product image. It turns that visual intent into shoppable pieces, store options, prices, resale paths, and try-on.

Search in your browser. The app is available for iPhone.

See a real app example

Start with a photo

Remode detects pieces in an outfit photo, including wide-leg brown trousers.

Find a similar piece

App result for brown trousers labeled Similar Match, with a link to the store.

Actual app screens. This result is labeled Similar Match. Prices and availability can change.

What makes it useful

The assistant is valuable when normal search words fail: you do not know the brand, the item is in a video, or the post has no product tag.

Image-first search

Start with the visual evidence instead of guessing keywords for the garment.

Shopping context

Use retailer, price, match quality, and second-hand signals before opening a store.

Return path

Save the result in Remode so the search can become price watch, try-on, or a later decision.

Best use cases

Remode is strongest when the user already has visual fashion intent.

Use Remode with AI assistants

Remode has been submitted to ChatGPT Apps and is available to AI assistants today over MCP at mcp.remode.app — assistants can call the Remode tool directly.

What buyer-side means in practice

Most shopping help online is paid for by the seller. A store's recommendation widget, a marketplace's sponsored row, a creator's affiliate link under a video: each of them earns when you buy a particular thing, so each has a reason to nudge you toward it. A buyer-side assistant is built the other way around. Its job is to answer your question, which in fashion is usually a short list: is this the piece, where is it cheaper, and should I buy it now.

Remode is designed as that kind of assistant. It reads the image you already have, identifies each fashion piece in it, and lays out stores, prices, and second-hand listings without letting a retailer pay for position. Ranking is never influenced by commercial terms, and the affiliate links some retailers provide do not move a result up the list. That separation matters in fashion specifically, because the same shirt can sit in many stores at different prices, and the store paying a commission is not necessarily the one that ships to you or has your size. The wider approach is explained on the reverse image search for clothes guide.

A full session, from saved image to decision

  1. Capture the reference. On iPhone, screenshot the post, pause a video on a clear frame and screenshot that, or save the product image from a retailer page to your camera roll.
  2. Open Remode and upload the screenshot. Remode detects every fashion piece in the frame, from clothing and shoes to bags, jewelry, and watches; tap the one you want to ask about. Each piece you search counts as one search.
  3. Wait for the result. A search takes about 5 to 15 seconds, and the answer arrives with a label: exact, best visual match, or similar.
  4. Read the store list as a comparison rather than an instruction. Retailer name, price, and whether the store ships to your region sit next to each other, and second-hand listings appear when Remode finds them.
  5. If the price is not right today, save the piece instead of buying it. Remode keeps watching it in the background for price drops, restocks, and second-hand options, for free, and watching does not use a search.
  6. For supported items, run the AI try-on (a Pro feature) before you commit, so the decision includes how the piece sits on you and not only what it costs.
  7. Open the store only when the label, the price, and the shipping region all line up.

How to read what the assistant gives back

The label is the first thing to read, because it tells you what kind of question the rest of the card can answer. With an exact result, the store comparison is about the same product, so price, shipping region, and condition are the whole decision. With a best visual match, Remode is telling you it found a close listing but is not confident it is the same product; treat the price as a reference point and confirm the details on the product page yourself. With a similar result you are comparing alternatives, and the question shifts from price hunting to whether this version does the job. Second-hand listings need one extra check, the seller's condition notes and return terms. An assistant that pretends every result is exact would save you a step and cost you a return; Remode's stance is the reverse. If it is not sure, it says so.

Habits that make the assistant more useful

  • Ask about one piece at a time. An outfit screenshot holds several products, and each has its own stores and prices; pick a piece, get the answer, then go back for the next.
  • Keep the original screenshot rather than a re-shared copy; every repost adds compression that erases the stitching and logo details an exact match depends on.
  • Save more than you buy. A saved piece costs nothing to watch, and a price drop or restock alert is a better trigger for a purchase than a countdown banner in a store.
  • Use the region routes. A lower price at a retailer that does not ship to you is not a lower price.
  • Look at the second-hand listings against the new price before deciding that a piece is out of budget.
  • If you already work inside an AI assistant, it can call Remode's MCP tool, find_fashion_items, with your image and get the same labeled results; the ChatGPT app has been submitted and is pending approval.

Where the assistant stops

Remode answers questions that start from a visual. It does not write outfit advice, rate a look, or tell you what is trending; if your question is styling rather than sourcing, it will have little to say. It also cannot surface products that have no online listing: a custom piece, a vintage one-off, or a small label that sells through direct messages will come back labeled similar, not exact. Remode does not sell anything itself, so checkout, returns, and sizing belong to the retailer or marketplace where you buy. The app is iOS only today; Android users can join a waitlist, and anyone can try 3 free photo searches in a browser without an account. When you need to confirm a result is the same product, the exact match guide covers what to check.

Common questions

What is an AI fashion shopping assistant?

It is a shopping tool that starts from what you saw, such as a photo or screenshot, then helps identify pieces and compare buying options.

Can Remode search from social screenshots?

Yes. Remode is built for screenshots and saved images from Instagram, TikTok, Pinterest, retailer pages, and your camera roll.

What can Remode do after finding an item?

You can compare retailers and prices, check second-hand options, save the item, watch for price changes, and use try-on when available.

Does an AI shopping assistant push me toward certain stores?

A seller-side tool often does, because it earns when you buy from a specific store. Remode's ranking is never influenced by commercial terms. Some links are affiliate links, and as an Amazon Associate Remode earns from qualifying purchases, but that does not change the order or the label of any result.

Can I use Remode inside ChatGPT or another AI assistant?

Remode has been submitted as a ChatGPT app and is waiting for approval. Separately, AI assistants that support MCP can already call the Remode tool find_fashion_items, pass it an image or a supported link, and receive the same exact, best visual match, and similar labels the app shows.

Does watching a saved item use up my free searches?

No. Watching is free and does not count as a search. Remode checks the saved piece in the background for price drops, restocks, and second-hand listings and alerts you when something changes, so waiting becomes a decision you can act on instead of a tab you forget.