Tutorial: Create a minimal API with ASP.
NET
Core
08/21/2024
By Rick Anderson and Tom Dykstra
Minimal APIs are architected to create HTTP APIs with minimal dependencies. They're ideal for
microservices and apps that want to include only the minimum files, features, and
dependencies in ASP.NET Core.
This tutorial teaches the basics of building a minimal API with ASP.NET Core. Another approach
to creating APIs in ASP.NET Core is to use controllers. For help with choosing between minimal
APIs and controller-based APIs, see APIs overview. For a tutorial on creating an API project
based on controllers that contains more features, see Create a web API.
Overview
This tutorial creates the following API:
ﾉ Expand table
API Description Request body Response body
GET /todoitems Get all to-do items None Array of to-do items
GET /todoitems/complete Get completed to-do items None Array of to-do items
GET /todoitems/{id} Get an item by ID None To-do item
POST /todoitems Add a new item To-do item To-do item
PUT /todoitems/{id} Update an existing item To-do item None
DELETE /todoitems/{id} Delete an item None None
Prerequisites
Visual Studio
Visual Studio 2022 with the ASP.NET and web development workload.
Create an API project
Visual Studio
Start Visual Studio 2022 and select Create a new project.
In the Create a new project dialog:
Enter Empty in the Search for templates search box.
Select the ASP.NET Core Empty template and select Next.
Name the project TodoApi and select Next.
In the Additional information dialog:
Select .NET 9.0
Uncheck Do not use top-level statements
Select Create
Examine the code
The Program.cs file contains the following code:
C#
var builder = WebApplication.CreateBuilder(args);
var app = builder.Build();
app.MapGet("/", () => "Hello World!");
app.Run();
The preceding code:
Creates a WebApplicationBuilder and a WebApplication with preconfigured defaults.
Creates an HTTP GET endpoint / that returns Hello World! :
Run the app
Visual Studio
Press Ctrl+F5 to run without the debugger.
Visual Studio displays the following dialog:
Select Yes if you trust the IIS Express SSL certificate.
The following dialog is displayed:
Select Yes if you agree to trust the development certificate.
For information on trusting the Firefox browser, see Firefox
SEC_ERROR_INADEQUATE_KEY_USAGE certificate error.
Visual Studio launches the Kestrel web server and opens a browser window.
Hello World! is displayed in the browser. The Program.cs file contains a minimal but
complete app.
Close the browser window.
Add NuGet packages
NuGet packages must be added to support the database and diagnostics used in this tutorial.
Visual Studio
From the Tools menu, select NuGet Package Manager > Manage NuGet Packages
for Solution.
Select the Browse tab.
Select Include Prelease.
Enter Microsoft.EntityFrameworkCore.InMemory in the search box, and then select
Microsoft.EntityFrameworkCore.InMemory .
Select the Project checkbox in the right pane and then select Install.
Follow the preceding instructions to add the
Microsoft.AspNetCore.Diagnostics.EntityFrameworkCore package.
The model and database context classes
In the project folder, create a file named Todo.cs with the following code:
C#
public class Todo
{
public int Id { get; set; }
public string? Name { get; set; }
public bool IsComplete { get; set; }
}
The preceding code creates the model for this app. A model is a class that represents data that
the app manages.
Create a file named TodoDb.cs with the following code:
C#
using Microsoft.EntityFrameworkCore;
class TodoDb : DbContext
{
public TodoDb(DbContextOptions<TodoDb> options)
: base(options) { }
public DbSet<Todo> Todos => Set<Todo>();
}
The preceding code defines the database context, which is the main class that coordinates
Entity Framework functionality for a data model. This class derives from the
Microsoft.EntityFrameworkCore.DbContext class.
Add the API code
Replace the contents of the Program.cs file with the following code:
C#
using Microsoft.EntityFrameworkCore;
var builder = WebApplication.CreateBuilder(args);
builder.Services.AddDbContext<TodoDb>(opt => opt.UseInMemoryDatabase("TodoList"));
builder.Services.AddDatabaseDeveloperPageExceptionFilter();
var app = builder.Build();
app.MapGet("/todoitems", async (TodoDb db) =>
await db.Todos.ToListAsync());
app.MapGet("/todoitems/complete", async (TodoDb db) =>
await db.Todos.Where(t => t.IsComplete).ToListAsync());
app.MapGet("/todoitems/{id}", async (int id, TodoDb db) =>
await db.Todos.FindAsync(id)
is Todo todo
? Results.Ok(todo)
: Results.NotFound());
app.MapPost("/todoitems", async (Todo todo, TodoDb db) =>
{
db.Todos.Add(todo);
await db.SaveChangesAsync();
return Results.Created($"/todoitems/{todo.Id}", todo);
});
app.MapPut("/todoitems/{id}", async (int id, Todo inputTodo, TodoDb db) =>
{
var todo = await db.Todos.FindAsync(id);
if (todo is null) return Results.NotFound();
todo.Name = inputTodo.Name;
todo.IsComplete = inputTodo.IsComplete;
await db.SaveChangesAsync();
return Results.NoContent();
});
app.MapDelete("/todoitems/{id}", async (int id, TodoDb db) =>
{
if (await db.Todos.FindAsync(id) is Todo todo)
{
db.Todos.Remove(todo);
await db.SaveChangesAsync();
return Results.NoContent();
}
return Results.NotFound();
});
app.Run();
The following highlighted code adds the database context to the dependency injection (DI)
container and enables displaying database-related exceptions:
C#
var builder = WebApplication.CreateBuilder(args);
builder.Services.AddDbContext<TodoDb>(opt => opt.UseInMemoryDatabase("TodoList"));
builder.Services.AddDatabaseDeveloperPageExceptionFilter();
var app = builder.Build();
The DI container provides access to the database context and other services.
Visual Studio
This tutorial uses Endpoints Explorer and .http files to test the API.
Test posting data
The following code in Program.cs creates an HTTP POST endpoint /todoitems that adds data
to the in-memory database:
C#
app.MapPost("/todoitems", async (Todo todo, TodoDb db) =>
{
db.Todos.Add(todo);
await db.SaveChangesAsync();
return Results.Created($"/todoitems/{todo.Id}", todo);
});
Run the app. The browser displays a 404 error because there's no longer a / endpoint.
The POST endpoint will be used to add data to the app.
Visual Studio
Select View > Other Windows > Endpoints Explorer.
Right-click the POST endpoint and select Generate request.
A new file is created in the project folder named TodoApi.http , with contents similar
to the following example:
HTTP
@TodoApi_HostAddress = https://localhost:7031
POST {{TodoApi_HostAddress}}/todoitems
###
The first line creates a variable that is used for all of the endpoints.
The next line defines a POST request.
The triple hashtag ( ### ) line is a request delimiter: what comes after it is for a
different request.
The POST request needs headers and a body. To define those parts of the request,
add the following lines immediately after the POST request line:
Content-Type: application/json
{
"name":"walk dog",
"isComplete":true
}
The preceding code adds a Content-Type header and a JSON request body. The
TodoApi.http file should now look like the following example, but with your port
number:
HTTP
@TodoApi_HostAddress = https://localhost:7057
POST {{TodoApi_HostAddress}}/todoitems
Content-Type: application/json
{
"name":"walk dog",
"isComplete":true
}
###
Run the app.
Select the Send request link that is above the POST request line.
The POST request is sent to the app and the response is displayed in the Response
pane.
Examine the GET endpoints
The sample app implements several GET endpoints by calling MapGet :
ﾉ Expand table
API Description Request body Response body
GET /todoitems Get all to-do items None Array of to-do items
GET /todoitems/complete Get all completed to-do items None Array of to-do items
GET /todoitems/{id} Get an item by ID None To-do item
C#
app.MapGet("/todoitems", async (TodoDb db) =>
await db.Todos.ToListAsync());
app.MapGet("/todoitems/complete", async (TodoDb db) =>
await db.Todos.Where(t => t.IsComplete).ToListAsync());
app.MapGet("/todoitems/{id}", async (int id, TodoDb db) =>
await db.Todos.FindAsync(id)
is Todo todo
? Results.Ok(todo)
: Results.NotFound());
Test the GET endpoints
Visual Studio
Test the app by calling the GET endpoints from a browser or by using Endpoints Explorer.
The following steps are for Endpoints Explorer.
In Endpoints Explorer, right-click the first GET endpoint, and select Generate
request.
The following content is added to the TodoApi.http file:
HTTP
GET {{TodoApi_HostAddress}}/todoitems
###
Select the Send request link that is above the new GET request line.
The GET request is sent to the app and the response is displayed in the Response
pane.
The response body is similar to the following JSON:
JSON
[
{
"id": 1,
"name": "walk dog",
"isComplete": true
}
]
In Endpoints Explorer, right-click the /todoitems/{id} GET endpoint and select
Generate request. The following content is added to the TodoApi.http file:
HTTP
GET {{TodoApi_HostAddress}}/todoitems/{id}
###
Replace {id} with 1 .
Select the Send request link that is above the new GET request line.
The GET request is sent to the app and the response is displayed in the Response
pane.
The response body is similar to the following JSON:
JSON
{
"id": 1,
"name": "walk dog",
"isComplete": true
}
This app uses an in-memory database. If the app is restarted, the GET request doesn't return
any data. If no data is returned, POST data to the app and try the GET request again.
Return values
ASP.NET Core automatically serializes the object to JSON and writes the JSON into the body
of the response message. The response code for this return type is 200 OK , assuming there
are no unhandled exceptions. Unhandled exceptions are translated into 5xx errors.
The return types can represent a wide range of HTTP status codes. For example, GET
/todoitems/{id} can return two different status values:
If no item matches the requested ID, the method returns a 404 status NotFound error
code.
Otherwise, the method returns 200 with a JSON response body. Returning item results in
an HTTP 200 response.
Examine the PUT endpoint
The sample app implements a single PUT endpoint using MapPut :
C#
app.MapPut("/todoitems/{id}", async (int id, Todo inputTodo, TodoDb db) =>
{
var todo = await db.Todos.FindAsync(id);
if (todo is null) return Results.NotFound();
todo.Name = inputTodo.Name;
todo.IsComplete = inputTodo.IsComplete;
await db.SaveChangesAsync();
return Results.NoContent();
});
This method is similar to the MapPost method, except it uses HTTP PUT. A successful response
returns 204 (No Content) . According to the HTTP specification, a PUT request requires the
client to send the entire updated entity, not just the changes. To support partial updates, use
HTTP PATCH.
Test the PUT endpoint
This sample uses an in-memory database that must be initialized each time the app is started.
There must be an item in the database before you make a PUT call. Call GET to ensure there's
an item in the database before making a PUT call.
Update the to-do item that has Id = 1 and set its name to "feed fish" .
Visual Studio
In Endpoints Explorer, right-click the PUT endpoint, and select Generate request.
The following content is added to the TodoApi.http file:
HTTP
PUT {{TodoApi_HostAddress}}/todoitems/{id}
###
In the PUT request line, replace {id} with 1 .
Add the following lines immediately after the PUT request line:
HTTP
Content-Type: application/json
{
"id": 1,
"name": "feed fish",
"isComplete": false
}
The preceding code adds a Content-Type header and a JSON request body.
Select the Send request link that is above the new PUT request line.
The PUT request is sent to the app and the response is displayed in the Response
pane. The response body is empty, and the status code is 204.
Examine and test the DELETE endpoint
The sample app implements a single DELETE endpoint using MapDelete :
C#
app.MapDelete("/todoitems/{id}", async (int id, TodoDb db) =>
{
if (await db.Todos.FindAsync(id) is Todo todo)
{
db.Todos.Remove(todo);
await db.SaveChangesAsync();
return Results.NoContent();
}
return Results.NotFound();
});
Visual Studio
In Endpoints Explorer, right-click the DELETE endpoint and select Generate request.
A DELETE request is added to TodoApi.http .
Replace {id} in the DELETE request line with 1 . The DELETE request should look like
the following example:
HTTP
DELETE {{TodoApi_HostAddress}}/todoitems/1
###
Select the Send request link for the DELETE request.
The DELETE request is sent to the app and the response is displayed in the Response
pane. The response body is empty, and the status code is 204.
Use the MapGroup API
The sample app code repeats the todoitems URL prefix each time it sets up an endpoint. APIs
often have groups of endpoints with a common URL prefix, and the MapGroup method is
available to help organize such groups. It reduces repetitive code and allows for customizing
entire groups of endpoints with a single call to methods like RequireAuthorization and
WithMetadata.
Replace the contents of Program.cs with the following code:
Visual Studio
C#
using Microsoft.EntityFrameworkCore;
var builder = WebApplication.CreateBuilder(args);
builder.Services.AddDbContext<TodoDb>(opt =>
opt.UseInMemoryDatabase("TodoList"));
builder.Services.AddDatabaseDeveloperPageExceptionFilter();
var app = builder.Build();
var todoItems = app.MapGroup("/todoitems");
todoItems.MapGet("/", async (TodoDb db) =>
await db.Todos.ToListAsync());
todoItems.MapGet("/complete", async (TodoDb db) =>
await db.Todos.Where(t => t.IsComplete).ToListAsync());
todoItems.MapGet("/{id}", async (int id, TodoDb db) =>
await db.Todos.FindAsync(id)
is Todo todo
? Results.Ok(todo)
: Results.NotFound());
todoItems.MapPost("/", async (Todo todo, TodoDb db) =>
{
db.Todos.Add(todo);
await db.SaveChangesAsync();
return Results.Created($"/todoitems/{todo.Id}", todo);
});
todoItems.MapPut("/{id}", async (int id, Todo inputTodo, TodoDb db) =>
{
var todo = await db.Todos.FindAsync(id);
if (todo is null) return Results.NotFound();
todo.Name = inputTodo.Name;
todo.IsComplete = inputTodo.IsComplete;
await db.SaveChangesAsync();
return Results.NoContent();
});
todoItems.MapDelete("/{id}", async (int id, TodoDb db) =>
{
if (await db.Todos.FindAsync(id) is Todo todo)
{
db.Todos.Remove(todo);
await db.SaveChangesAsync();
return Results.NoContent();
}
return Results.NotFound();
});
app.Run();
The preceding code has the following changes:
Adds var todoItems = app.MapGroup("/todoitems"); to set up the group using the URL
prefix /todoitems .
Changes all the app.Map<HttpVerb> methods to todoItems.Map<HttpVerb> .
Removes the URL prefix /todoitems from the Map<HttpVerb> method calls.
Test the endpoints to verify that they work the same.
Use the TypedResults API
Returning TypedResults rather than Results has several advantages, including testability and
automatically returning the response type metadata for OpenAPI to describe the endpoint. For
more information, see TypedResults vs Results.
The Map<HttpVerb> methods can call route handler methods instead of using lambdas. To see
an example, update Program.cs with the following code:
Visual Studio
C#
using Microsoft.EntityFrameworkCore;
var builder = WebApplication.CreateBuilder(args);
builder.Services.AddDbContext<TodoDb>(opt =>
opt.UseInMemoryDatabase("TodoList"));
builder.Services.AddDatabaseDeveloperPageExceptionFilter();
var app = builder.Build();
var todoItems = app.MapGroup("/todoitems");
todoItems.MapGet("/", GetAllTodos);
todoItems.MapGet("/complete", GetCompleteTodos);
todoItems.MapGet("/{id}", GetTodo);
todoItems.MapPost("/", CreateTodo);
todoItems.MapPut("/{id}", UpdateTodo);
todoItems.MapDelete("/{id}", DeleteTodo);
app.Run();
static async Task<IResult> GetAllTodos(TodoDb db)
{
return TypedResults.Ok(await db.Todos.ToArrayAsync());
}
static async Task<IResult> GetCompleteTodos(TodoDb db)
{
return TypedResults.Ok(await db.Todos.Where(t =>
t.IsComplete).ToListAsync());
}
static async Task<IResult> GetTodo(int id, TodoDb db)
{
return await db.Todos.FindAsync(id)
is Todo todo
? TypedResults.Ok(todo)
: TypedResults.NotFound();
}
static async Task<IResult> CreateTodo(Todo todo, TodoDb db)
{
db.Todos.Add(todo);
await db.SaveChangesAsync();
return TypedResults.Created($"/todoitems/{todo.Id}", todo);
}
static async Task<IResult> UpdateTodo(int id, Todo inputTodo, TodoDb db)
{
var todo = await db.Todos.FindAsync(id);
if (todo is null) return TypedResults.NotFound();
todo.Name = inputTodo.Name;
todo.IsComplete = inputTodo.IsComplete;
await db.SaveChangesAsync();
return TypedResults.NoContent();
}
static async Task<IResult> DeleteTodo(int id, TodoDb db)
{
if (await db.Todos.FindAsync(id) is Todo todo)
{
db.Todos.Remove(todo);
await db.SaveChangesAsync();
return TypedResults.NoContent();
}
return TypedResults.NotFound();
}
The Map<HttpVerb> code now calls methods instead of lambdas:
C#
var todoItems = app.MapGroup("/todoitems");
todoItems.MapGet("/", GetAllTodos);
todoItems.MapGet("/complete", GetCompleteTodos);
todoItems.MapGet("/{id}", GetTodo);
todoItems.MapPost("/", CreateTodo);
todoItems.MapPut("/{id}", UpdateTodo);
todoItems.MapDelete("/{id}", DeleteTodo);
These methods return objects that implement IResult and are defined by TypedResults:
C#
static async Task<IResult> GetAllTodos(TodoDb db)
{
return TypedResults.Ok(await db.Todos.ToArrayAsync());
}
static async Task<IResult> GetCompleteTodos(TodoDb db)
{
return TypedResults.Ok(await db.Todos.Where(t => t.IsComplete).ToListAsync());
}
static async Task<IResult> GetTodo(int id, TodoDb db)
{
return await db.Todos.FindAsync(id)
is Todo todo
? TypedResults.Ok(todo)
: TypedResults.NotFound();
}
static async Task<IResult> CreateTodo(Todo todo, TodoDb db)
{
db.Todos.Add(todo);
await db.SaveChangesAsync();
return TypedResults.Created($"/todoitems/{todo.Id}", todo);
}
static async Task<IResult> UpdateTodo(int id, Todo inputTodo, TodoDb db)
{
var todo = await db.Todos.FindAsync(id);
if (todo is null) return TypedResults.NotFound();
todo.Name = inputTodo.Name;
todo.IsComplete = inputTodo.IsComplete;
await db.SaveChangesAsync();
return TypedResults.NoContent();
}
static async Task<IResult> DeleteTodo(int id, TodoDb db)
{
if (await db.Todos.FindAsync(id) is Todo todo)
{
db.Todos.Remove(todo);
await db.SaveChangesAsync();
return TypedResults.NoContent();
}
return TypedResults.NotFound();
}
Unit tests can call these methods and test that they return the correct type. For example, if the
method is GetAllTodos :
C#
static async Task<IResult> GetAllTodos(TodoDb db)
{
return TypedResults.Ok(await db.Todos.ToArrayAsync());
}
Unit test code can verify that an object of type Ok<Todo[]> is returned from the handler
method. For example:
C#
public async Task GetAllTodos_ReturnsOkOfTodosResult()
{
// Arrange
var db = CreateDbContext();
// Act
var result = await TodosApi.GetAllTodos(db);
// Assert: Check for the correct returned type
Assert.IsType<Ok<Todo[]>>(result);
}
Prevent over-posting
Currently the sample app exposes the entire Todo object. In production applications, a subset
of the model is often used to restrict the data that can be input and returned. There are
multiple reasons behind this and security is a major one. The subset of a model is usually
referred to as a Data Transfer Object (DTO), input model, or view model. DTO is used in this
article.
A DTO can be used to:
Prevent over-posting.
Hide properties that clients aren't supposed to view.
Omit some properties to reduce payload size.
Flatten object graphs that contain nested objects. Flattened object graphs can be more
convenient for clients.
To demonstrate the DTO approach, update the Todo class to include a secret field:
C#
public class Todo
{
public int Id { get; set; }
public string? Name { get; set; }
public bool IsComplete { get; set; }
public string? Secret { get; set; }
}
The secret field needs to be hidden from this app, but an administrative app could choose to
expose it.
Verify you can post and get the secret field.
Create a file named TodoItemDTO.cs with the following code:
C#
public class TodoItemDTO
{
public int Id { get; set; }
public string? Name { get; set; }
public bool IsComplete { get; set; }
public TodoItemDTO() { }
public TodoItemDTO(Todo todoItem) =>
(Id, Name, IsComplete) = (todoItem.Id, todoItem.Name, todoItem.IsComplete);
}
Replace the contents of the Program.cs file with the following code to use this DTO model:
Visual Studio
C#
using Microsoft.EntityFrameworkCore;
var builder = WebApplication.CreateBuilder(args);
builder.Services.AddDbContext<TodoDb>(opt =>
opt.UseInMemoryDatabase("TodoList"));
builder.Services.AddDatabaseDeveloperPageExceptionFilter();
var app = builder.Build();
RouteGroupBuilder todoItems = app.MapGroup("/todoitems");
todoItems.MapGet("/", GetAllTodos);
todoItems.MapGet("/complete", GetCompleteTodos);
todoItems.MapGet("/{id}", GetTodo);
todoItems.MapPost("/", CreateTodo);
todoItems.MapPut("/{id}", UpdateTodo);
todoItems.MapDelete("/{id}", DeleteTodo);
app.Run();
static async Task<IResult> GetAllTodos(TodoDb db)
{
return TypedResults.Ok(await db.Todos.Select(x => new
TodoItemDTO(x)).ToArrayAsync());
}
static async Task<IResult> GetCompleteTodos(TodoDb db) {
return TypedResults.Ok(await db.Todos.Where(t => t.IsComplete).Select(x =>
new TodoItemDTO(x)).ToListAsync());
}
static async Task<IResult> GetTodo(int id, TodoDb db)
{
return await db.Todos.FindAsync(id)
is Todo todo
? TypedResults.Ok(new TodoItemDTO(todo))
: TypedResults.NotFound();
}
static async Task<IResult> CreateTodo(TodoItemDTO todoItemDTO, TodoDb db)
{
var todoItem = new Todo
{
IsComplete = todoItemDTO.IsComplete,
Name = todoItemDTO.Name
};
db.Todos.Add(todoItem);
await db.SaveChangesAsync();
todoItemDTO = new TodoItemDTO(todoItem);
return TypedResults.Created($"/todoitems/{todoItem.Id}", todoItemDTO);
}
static async Task<IResult> UpdateTodo(int id, TodoItemDTO todoItemDTO, TodoDb
db)
{
var todo = await db.Todos.FindAsync(id);
if (todo is null) return TypedResults.NotFound();
todo.Name = todoItemDTO.Name;
todo.IsComplete = todoItemDTO.IsComplete;
await db.SaveChangesAsync();
return TypedResults.NoContent();
}
static async Task<IResult> DeleteTodo(int id, TodoDb db)
{
if (await db.Todos.FindAsync(id) is Todo todo)
{
db.Todos.Remove(todo);
await db.SaveChangesAsync();
return TypedResults.NoContent();
}
return TypedResults.NotFound();
}
Verify you can post and get all fields except the secret field.
Troubleshooting with the completed sample
If you run into a problem you can't resolve, compare your code to the completed project. View
or download completed project (how to download).
Next steps
Configure JSON serialization options.
Handle errors and exceptions: The developer exception page is enabled by default in the
development environment for minimal API apps. For information about how to handle
errors and exceptions, see Handle errors in ASP.NET Core APIs.
For an example of testing a minimal API app, see this GitHub sample .
OpenAPI support in minimal APIs.
Quickstart: Publish to Azure.
Organizing ASP.NET Core Minimal APIs .
Learn more
See Minimal APIs quick reference
