# Checkout Bench

> Checkout Bench runs AI shopping agents through real, signed-in store checkouts in fresh cloud browsers and grades the final cart in code. Ordinary tests stop at the quote and put the cart back. Retailers can request a private test of their store, the test suite at /lab runs assistants against real quotes, and Nook's agent API is documented under /docs.

## Pages

- [Home](https://www.checkoutbench.ai/): what exists today and five directions ([markdown](https://www.checkoutbench.ai/index.md))
- [Test suite](https://www.checkoutbench.ai/suite): recent real Nook runs and the 30-run Test flight campaign (scrubbed), plus a leaderboard and task bank that are example data ([markdown](https://www.checkoutbench.ai/suite.md))
- [Lab](https://www.checkoutbench.ai/lab/): the test suite itself; sign in to run assistants against real quotes (not indexed)

## Nook docs

> As of Nov 3, 2026, 2:45 PM PST, from Nook main 680daa2f.

- [Overview](https://www.checkoutbench.ai/docs) ([markdown](https://www.checkoutbench.ai/docs/index.md))
- [Supported stores](https://www.checkoutbench.ai/docs/stores) ([markdown](https://www.checkoutbench.ai/docs/stores.md))
- [How a request flows](https://www.checkoutbench.ai/docs/request-flow) ([markdown](https://www.checkoutbench.ai/docs/request-flow.md))
- [API basics](https://www.checkoutbench.ai/docs/api) ([markdown](https://www.checkoutbench.ai/docs/api.md))
- [Testing lab](https://www.checkoutbench.ai/docs/lab) ([markdown](https://www.checkoutbench.ai/docs/lab.md))
- [Product identity rules](https://www.checkoutbench.ai/docs/product-identity) ([markdown](https://www.checkoutbench.ai/docs/product-identity.md))
- [Known limits and open issues](https://www.checkoutbench.ai/docs/limits) ([markdown](https://www.checkoutbench.ai/docs/limits.md))
- [Changelog](https://www.checkoutbench.ai/docs/changelog) ([markdown](https://www.checkoutbench.ai/docs/changelog.md))


## Facts

- 40: stores and sites a person can connect an account to today: 12 built in, 28 set up from their web address
- 7: stores where a full cart and total come back today: Amazon, Uber Eats, Instacart, Walmart, Target, Chewy, DETOURE
- 1: fresh cloud browser per run, thrown away afterwards
- 384: trace events in one real run: DeepSeek V4.1 Flash asked Amazon for a DJI mic; Nook blocked all five quotes as product mismatch, nothing bought