Can an AI agent buy from your store?

Checkout Bench runs AI agents through real and simulated environments. Retailers automate AI checkout testing, gather data, and build proprietary models that perfectly fit their own checkout.

Track and Fix how AI agents react to your store in real time.

9:41
BatteryShack
Shack RewardsPower up.
Pay less.
Guest checkout
Ship to214 Market Street, San Francisco
DeliveryStandard, 2 to 3 days
Continue as guest
Free shipping on orders over $35.Easy returns within 30 days of delivery.
No account is needed to check out.
ChargersBatteriesEarbudsPower banks
Weekend Charge SaleUp to 30% off cables and power banks
Best seller
ShackCharge USB-C Cable, 2 m, braided
4.8 · 2,184 reviews
$12.99 $16.99
2 m1 m3 m
Add to cart
In stock and ready for pickup today at Shack #214.
Delivery Arrives Tuesday with free shipping.
Your cart
ShackCharge USB-C, 2 mQty 1 · Black$12.99
Ships free with Shack Rewards
Saved home address · saved card
Go to checkout
Checkout
Item subtotal$12.99
ShippingFree
Estimated tax$1.14
Order total$14.13
Pickup Shack #214, today after 2 PMPayment Card on file
Place order · $14.13
0:00
0.0s

Checkout Bench provides easy-to-use simulators to see how any AI agent or model would navigate through your stores and apps today.

Submit your store

During our call, we'll onboard you to Checkout Lab and the Checkout Bench API.

The deepest benchmark in agentic commerce.

Most Agent-Ready Platform

2026 models
Amazon

Amazon is the easiest major store for AI agents to navigate across the major 2026 models.

Frontier AI research meets business owners where they are.

Read the research

See exactly where ChatGPT, Claude, Gemini and more struggle with navigating your site.

See exactly where ChatGPT, Claude, Gemini and more struggle with navigating your site.

Explore the network traces down to hundredths of a second to see where AIs do and don't understand your business.

1Submit your store and start testing within 24 hours.

Submit your store and start testing within 24 hours.

StoreTarget
ModelGPT-5.6 Luna
TaskA 2 m USB-C cable
Budget$30 cap

Send GPT-5.6 Luna to Target to buy a 2 m USB-C cable, keeping to a $30 cap, and score how far it gets toward a ready checkout.

Deploy any test to mimic your customers: different models, products, budgets, and more.

2Automate thousands of AI agent runs in one place.

Automate thousands of AI agent runs in one place.

Sign inNook
  • browser_ready
  • signed_in
  • search_results
  • product_chosen
product_choice_requires_user

Two colors matched the case, so the agent asked the person and carried on with “Lime”.

CartShopify
  • product_page
  • added_to_cart
  • cart_checkout
Address and identityPlaid
  • address_selected

The saved home address was already the default.

PaymentLink by Stripe
  • payment_selected

The saved card was already the default.

CheckoutCheckout Bench
  • checkout_read
  • quote_checked
  • quote_published

Item subtotal $351.20 sits inside the ±$6.98 allowance around $349.00.

Ready Checkout

Follow each run from store entry to checkout, with actions and results organized by stage.

3Run every checkout in the lab.

Developers

Run every checkout in the lab.

Steps run in parallel, the slower environments get cancelled, and every run is traced.

Run Graphidle
task_intakequeued
member_sessionqueued
cart_resolverqueued
live_storefrontqueued
sim_replicaqueued
snapshot_mirrorqueued
agent_rolloutqueued
receipt_verifierqueued
trace_recorderqueued
graded_receiptqueued
orchestrator.trace()
    0 events captured
    1. Idle
    2. Intake
    3. Prepare
    4. Race
    5. Handoff
    6. Cancel
    7. Rollout
    8. Verify
    9. Record
    10. Done
    4Simulate different context and preference setups.

    Simulate different context and preference setups.

    A person0.0 s
    1. Search
    2. Color
    3. Cart
    4. Address
    5. Review

    Typed a search and scrolled the results

    An AI agent0.0 s
    1. Search
    2. Color
    3. Cart
    4. Address
    5. Review

    Read 24 results, 3 matched the brief

    Same order, same total Order review $351.20 Wireless headphones · Lime

    Agents now shop on behalf of people. We run the same order the way a person would and the way an agent does. When both reach the same order review with the same total, your store is ready for both.

    5Test your AI store performance against competitors.

    Test your AI store performance against competitors.

    Checkout requests by store

    Ready versus blocked checkout requests at leading retailers.

    Anonymized
    Ready CheckoutBlockedStill preparing
    Amazon
    Scorecard
    323
    Ready Checkouts from 479 requests
    Median time
    40.6 s
    Checkout total
    $8.70
    Uber Eats
    Scorecard
    40
    Ready Checkouts from 159 requests
    Median time
    46.0 s
    Checkout total
    $26.79
    Instacart
    Scorecard
    24
    Ready Checkouts from 123 requests
    Median time
    40.4 s
    Checkout total
    $26.84
    Walmart
    Scorecard
    20
    Ready Checkouts from 68 requests
    Median time
    37.0 s
    Checkout total
    $8.80
    Chewy
    Scorecard
    10
    Ready Checkouts from 33 requests
    Median time
    61.2 s
    Checkout total
    $9.35
    Target
    Scorecard
    6
    Ready Checkouts from 42 requests
    Median time
    45.8 s
    Checkout total
    $31.52

    Every scorecard comes from AI agents walking a store's real checkout, signed in as a member, all the way to the final total.

    • Ready checkoutsHow many agent checkout requests reached a checkout ready to place.
    • Where agents stallThe stage where unfinished runs stopped, from store entry to payment.
    • Side by sideYour store next to the retailers you compete with, run the same way.

    See how your store stacks up against different retailers in your industry.

    6Refine guest checkouts and loyalty experiences with real data.

    Refine guest checkouts and loyalty experiences with real data.

    Amazon · GPT-5.6 Sol
    Opening a fresh browser
    0seconds
    Ready Checkouts this month
    184
    +1 each Ready Checkout
    Opening a fresh browserUber Eats · DeepSeek V4.1 Flash0s
    Opening a fresh browserInstacart · GPT-5.6 Luna0s
    Opening a fresh browserWalmart · Grok 4.70s
    7Deploy a swarm of agents to see if it reaches checkout.

    Deploy a swarm of agents to see if it reaches checkout.

    Anonymized
    GPT-5.6 SolGPT-5.6 Sol415 runs
    DeepSeek V4.1 FlashDeepSeek V4.1 Flash279 runs
    GPT-5.6 LunaGPT-5.6 Luna175 runs
    Grok 4.7Grok 4.720 runs
    Lab models
    Checkout Bench
    Amazon323
    Uber Eats40
    Instacart24
    Walmart20
    Chewy10
    Target6
    Ready Checkouts per store

    On Amazon, agents reached a ready checkout 323 times, in a median of 40.6 seconds.

    In the lab, 47% of agent checkout requests across six stores reached a ready checkout.

    8Custom training environments for thousands of sites

    Training data

    Custom training environments for thousands of sites

    Each store becomes an environment where agents learn to catch the wrong size, swap, add-on or fee before a cart is approved.

    Frontier AI research meets business owners where they are.

    Read the research

    Book a call
    with us

    “There should be as many models as there are companies in the world. Because after all, what is a company? A company is a learning system.”.”

    Satya NadellaSatya NadellaChairman and CEO, Microsoft

    We build your private test on this site.