Domino's ended fiscal 2025 with more than 22,100 stores in over 90 markets, 99 percent of them owned by independent franchisees, and global retail sales above $20.1 billion (Domino's Pizza, 23 February 2026). Franchisees set their own prices and pick their own local coupons, so the website shows you a price only once you have chosen a store, and the store you chose decides which coupons exist. That single design fact is why Domino's is a store-level dataset rather than a menu, and why the people who read it at scale (franchise consultants, competitors, price-intelligence firms, and a surprising number of hobbyist developers) end up asking us about proxies.
The hobbyists are the second reason this page exists. For years, community-written libraries on public package registries have wrapped the same store-finder, menu and coupon endpoints the Domino's website calls, which is what people mean by the "Domino's API". Those calls are metered per address exactly like a browser's. Below: Domino's as a target, the real jobs and the ones a proxy cannot do, the door on 24 August 2026 in numbers, then which type, how to set up, how to size, and where it ends.
Short answer
Residential proxies pinned to the metro of the stores you read, driving a real browser or, for the endpoint route, a client that behaves like one, holding one store per session. Logins belong on a static ISP proxy, one account each. Datacenter exits load the homepage (we checked), and the store endpoints behind it are where a shared address gets throttled.
Domino's as a target
Three layers. The store finder takes an address and returns the store that delivers to it (or the carryout stores nearby), which is a delivery-zone question as well as a location one. The store menu carries that store's prices, and the coupon page carries that store's local offers alongside the national ones. And the 90-plus international markets are run by master franchisees with their own websites, menus and currencies, so the Japanese, Australian or Indian Domino's is a different site under a different company.
The endpoints behind the US site are what the unofficial libraries wrap. They return the same data the pages render, which is convenient, and they are metered the same way, which is the part the libraries' READMEs rarely mention.
Jobs a proxy is right for
- Store-level price monitoring. The same large pepperoni across a metro, on a schedule. Franchise consultants benchmark operators against each other, competitors watch a market, price-intelligence firms build the dataset.
- Local coupon tracking. Which coupons each store runs, when they change, and how they differ across a city. Deal sites read it weekly; franchisees compare themselves to their neighbors.
- Delivery-zone discovery. Which store serves which address, and where the gaps are. Site selection for new stores and competitor mapping both start here.
- Location data. The store list with hours and services, across 22,100 stores, refreshed as stores open. The one dataset that grows by hundreds of stores a year (776 net new in 2025).
- International menu and price comparison. Same brand, different master franchisee, different site, read from an exit in that country.
Jobs a proxy is not for
Driver accounts. Identity checks, not addresses.
Rewards and multiple accounts. They attach to phone or email and the app on a device; farming offers runs against Domino's terms.
Orders you do not mean to pay for. Placing test orders through the unofficial libraries against real stores is not data collection, and a proxy does not make it something else.
The door on 24 August 2026
dominos.com handed our German datacenter server its real homepage on the first request, HTTP 200 through Akamai. The twenty free proxies from our free proxy list followed, a single request apiece: 15 failed to connect and 5 loaded the homepage. So the front door is open to nearly everyone, and the defense sits deeper: the store-finder, menu and coupon endpoints are the natural place for per-address metering, because a customer asks for one store and a script asks for five hundred. Nobody publishes that limit and we have not measured it; the first rising 403 or 429 rate per exit is the number, and backing off is the response. Scraping past Akamai covers the layer we saw at the door, and fixing 429 rate limits covers the response you will meet behind it.
Metro-pinned residential exit
same city as the store
Store chosen by address
sticky session
Menu + local coupons
record store id, time
Rotate
next store, fresh exit
Type by job
| Domino's job | Proxy type | Why |
|---|---|---|
| Store-level price monitoring | Rotating US residential, metro-pinned, sticky per store | Prices are per franchisee; each store is a session |
| Local coupon tracking | Rotating US residential, metro-pinned | Coupons are per store and change weekly |
| Delivery-zone discovery | Rotating US residential, sticky per address | The finder answers per address |
| Locator and location data | US residential, slow pacing | Paged queries from one address is the metered pattern |
| International sites | Residential in that country | Separate master franchisee, separate site |
| Any login | ISP (static residential) | One fixed, trusted location per account |
| One manual look at one store | Datacenter, or a fresh free proxy | The homepage answered; the sweep will not survive |
Background on why the two address types are treated differently: datacenter vs residential proxies.
Free proxies here
Five of twenty opened the door, which makes Domino's the friendliest target in this cluster for a one-off look and no friendlier for a sweep. Use a free proxy to test that your parser reads a store menu correctly, verify it in the proxy checker first, and read when free proxies are fine for the line. For the schedule, our US residential proxies start at $0.44/GB, pay as you go, no KYC, pinned to the metro.
Setup notes specific to Domino's
- Choose the store by address, then hold the session. Enter an address in the target neighborhood, accept the store the finder returns, read the menu and the coupon page, record the store number and the timestamp, then rotate. For the browser route, hand Playwright the proxy at launch as the Playwright proxy guide shows.
- If you use an unofficial library, give it the same manners. Route it through the same residential exit, keep its cookies, pace it like a person, and expect it to break when the site changes. It is a convenience, not a license.
- Diff coupons weekly, prices daily. Coupons rotate on a cycle; prices move rarely. Read each on its own schedule instead of re-reading everything.
- Retire an exit on the first block or 429. Retrying burns the address. Avoiding IP bans while scraping continues the list.
Sizing
One metro, 60 stores, prices daily and coupons weekly:
store sessions per day 60
pages per session 3-4 (menu, coupons, finder)
sessions per exit per day 4-6
exits in rotation ~12-15 US residential, metro-pinned
Bandwidth is the unit, these pages are small with images blocked, and the store count therefore sets the bill. A monthly benchmark costs a month under our pricing, with a balance that waits between runs.
Where it ends
Your session gets a local address from the exit, and nothing else. It does not make the unofficial libraries official, it does not touch driver or rewards accounts, and it does not make scraping compatible with Domino's terms, which prohibit it. Inside that, a metro-pinned residential exit lets a legitimate data job read public store prices and coupons the way customers of each store see them: one store per session, human pacing, and an eye on the block rate.
The chain that hides US prices inside its app instead of behind a store choice is in proxies for McDonald's; the marketplaces where every franchisee's delivery price is readable per address are in proxies for DoorDash.
Sources
- Domino's Pizza, fourth quarter and fiscal 2025 results, 23 February 2026 (more than 22,100 stores in over 90 markets; 99 percent franchisee-owned; global retail sales over $20.1 billion; 776 net store growth): prnewswire.com
- HProxy measurement, 24 August 2026: one direct request from a datacenter server (HTTP 200 through Akamai) and 20 elite HTTP proxies from our free proxy list (15 no connection, 5 homepage).