What are endpoints in RTK Query?
In RTK Query, endpoints define the operations your application can perform with an API — such as fetching, creating, updating, or deleting data. Each endpoint describes how a specific request should be made and how RTK Query should cache or update the response.
- 1
Query endpoints: Used for fetching and reading data from an API. Defined using
builder.query(). - 2
Mutation endpoints: Used for creating, updating, or deleting data. Defined using
builder.mutation().
- 1
Define the API routes and request configurations for each operation.
- 2
Automatically generate React hooks like
useGetUsersQueryoruseAddUserMutation. - 3
Control caching, invalidation, and refetching behavior for specific data sets.
- 4
Allow strong typing and automatic inference when using TypeScript.
Endpoints are the core building blocks of RTK Query APIs, defining how your application interacts with backend services in a clean, declarative way.
Scenario Questions
0-2 years experience
- 1You need to fetch a list of users using RTK Query. How would you define an endpoint for that, and what hook does RTK Query generate for you to use in a component?
- 2If you forget to add a `providesTags` field to a query endpoint, what impact does that have when you later add a mutation that should refresh that data?
2-5 years experience
- 1Your team added a new endpoint to fetch product details, but the UI never updates after a product‑edit mutation. Walk me through how you would debug the endpoint configuration and tag usage.
- 2We need to skip a query when a certain prop is false. How would you adjust the generated hook usage or the endpoint definition to conditionally avoid the request without adding extra local state?
5-8 years experience
- 1In a large application with dozens of endpoints across multiple feature slices, how would you organize endpoint definitions to avoid duplication and keep the generated hooks maintainable?
- 2Discuss the trade‑offs between enabling `refetchOnFocus` globally for an endpoint versus manually invalidating tags after a user returns to the app.
8+ years experience
- 1Our organization is migrating from a custom fetch layer to RTK Query across several micro‑frontends. What architectural considerations around endpoint design, versioning, and shared cache would you raise?
- 2How would you design a global strategy for handling endpoint failures—such as retries and exponential backoff—while still allowing individual endpoints to customize their error handling?
Follow-up Questions
- What is the naming convention for the hook generated from an endpoint?
- How does RTK Query know which slice of the cache to invalidate when a mutation runs?
- Can you change the generated hook name without renaming the endpoint?