Rappi restaurants, menus and grocery prices — by country and by city

Nine: Colombia, Mexico, Brazil, Argentina, Chile, Peru, Uruguay, Ecuador and Costa Rica. They are separate storefronts with separate catalogues and currencies, and the API treats them that way. On 2026-09-23 all nine answered with their own default city - Bogota, Ciudad de Mexico, Grande Sao Paulo, Buenos Aires, Santiago de Chile, Lima, Montevideo, Quito and San Jose - and their own restaurants. One consequence matters for your database: Rappi store ids are namespaced per country, not globally unique, so a low id such as 98 exists in more than one country as a different business. Key your records on country plus store_id, which is exactly how every record the API returns is labelled.

Yes, and it is the main reason to use this API rather than a single national scrape. On 2026-09-23 six Colombian cities returned six completely disjoint sets: across all 15 pairs of those cities, zero store ids were shared. The per-city restaurant counts Rappi itself publishes were Bogota 6,252, Medellin 4,739, Barranquilla 4,593, Cali 3,638, Pereira 1,415 and Cartagena 1,276. Each response also carries the city's own name and coordinates. Use the cities action to get the list a country actually serves - 34 for Colombia, 53 for Mexico, 53 for Brazil, 16 for Chile, 10 each for Peru and Ecuador, 9 for Argentina, 2 for Uruguay and 1 for Costa Rica - and a city outside that list returns NOT_FOUND rather than quietly answering with a different city's restaurants.

The whole menu, grouped into the sections the restaurant uses. Across the 10 Bogota restaurants measured on 2026-09-23 the API returned 742 items in total, ranging from 10 items for a small kitchen to 158 for a cafe, with 6 sections and 49 items for Home Burgers. Every item carries its price, its description, whether it is currently available, whether it has options or toppings to choose, and its image. Where an item is discounted you also get original_price and the discount percent; on that sample 5 of 742 items were discounted, so expect original_price to be null most of the time rather than assuming a promotion is missing.

Every restaurant page publishes its own machine-readable menu, and the API checks its parsed prices against it on every call and reports the result in price_cross_check. On 2026-09-23, across 742 menu items from 10 restaurants, every item matched. As a second and fully independent check, 6 of those restaurants were re-read from scratch and all 377 prices the API returned were found in the text the page itself prints. Prices are whole units of the local currency, exactly as the source publishes them - a 146000 in Colombia is 146,000 COP, printed on the page as $ 146.000 - so nothing is divided by 100. The currency is returned per restaurant from the source's own declaration rather than assumed from the country.

Yes. Pass cuisine and you get Rappi's own complete list for that cuisine in that city, not a page of it. Measured in Bogota on 2026-09-23 against an unfiltered city page of 30 rows: hamburguesa returned 58 restaurants, sushi 41, pizza 36, mexicana 32 and postres 131 - and every single returned row carried that cuisine's own id, 131 of 131 for postres and 32 of 32 for mexicana. The full cuisine vocabulary for a city comes back with any unfiltered search as cuisine_taxonomy, 32 cuisines in Bogota. One honest limitation: a cuisine-filtered page does not republish that vocabulary, so on those responses the rows carry cuisine_ids but not cuisine names, and the API tells you which case you are in with cuisine_names_available rather than inventing names.

store/search lists stores by store type and store/detail returns a store's record plus its product aisles with prices. In Colombia on 2026-09-23 the market type listed 68 stores and rappimall-parent 1,574. A store's products come back grouped into aisles with each item's price, pre-discount price, sale type and category - for example a convenience store returned 60 priced items across 4 aisles. Two honest limits: the grocery store listing is country-wide rather than city-scoped, and the API reports that as city_scoped false instead of pretending otherwise; and a store type Rappi does not publish returns NOT_FOUND rather than silently answering with the default type.

store/detail takes a pages parameter and it genuinely goes deeper. Measured on one Bogota convenience store on 2026-09-23: pages=1 returned 4 aisles and 60 products, pages=2 returned 13 aisles and 196 products, and pages=3 returned 25 aisles and 376 products. The default is 2. Every response also carries aisle_directory, the store's own list of aisles with the number of products Rappi says each one holds - for that store, Bebidas 424, Novedades 505 and Carnes 85 - so you can see what remains beyond the pages you asked for. This is the store's shelf catalogue as Rappi renders it, not a guaranteed complete inventory dump, and the directory is what tells you the difference.

No, and rather than offer a parameter that does nothing we left it out. We tested it properly: querying one store repeatedly with the same coordinates produced results that differed from each other as much as two different cities did, and coordinates in Madrid, Spain returned a Bogota store's identical shelves. The shelf content rotates between variants on its own and the address does not steer it, so there is no lat/lng parameter on store/detail. The store's own record still carries its real address and coordinates.

Individual customer review text - only the aggregate score and the 1-to-5 histogram are published, so there is no reviews list to return. Also absent: per-item stock counts, restaurant phone numbers, and courier tracking. On the 10-restaurant sample all stores carried name, rating, review count, address, coordinates, currency, delivery cost, cuisines, schedule, brand name, logo and image; the review histogram was present on 7 of 10, estimated delivery time on 7 of 10 and offers on 7 of 10. Anything the source does not publish comes back null rather than guessed.

Rappi API is a ReefAPI endpoint group for latin america's delivery marketplace: restaurants, menus with prices, and grocery stores across 9 countries. It returns live JSON through POST requests under /rappi/v1.

Yes. ReefAPI starts with 1,000 free credits, no card required. Rappi calls use the same shared credit balance as every other ReefAPI engine.

No login to Rappi is needed for the API response. You call ReefAPI with your x-api-key header, and the playground can run live examples before you create a production key.

The page example is captured from a live restaurant/search call, and production requests fetch live data through ReefAPI rather than a static sample.

Rappi actions currently cost 1-3 credits per successful call. Failed or blocked calls are free. All APIs draw from one credit pool.