Getting started

This chapter introduces some core concepts of mypy, including function annotations, the typing module, library stubs, and more.

Be sure to read this chapter carefully, as the rest of the documentation may not make much sense otherwise.

Installing and running mypy

Mypy requires Python 3.6 or later to run. Once you’ve installed Python 3, install mypy using pip:

$ python3 -m pip install mypy

Once mypy is installed, run it by using the mypy tool:

$ mypy

This command makes mypy type check your file and print out any errors it finds. Mypy will type check your code statically: this means that it will check for errors without ever running your code, just like a linter.

This means that you are always free to ignore the errors mypy reports and treat them as just warnings, if you so wish: mypy runs independently from Python itself.

However, if you try directly running mypy on your existing Python code, it will most likely report little to no errors: you must add type annotations to your code to take full advantage of mypy. See the section below for details.


Although you must install Python 3 to run mypy, mypy is fully capable of type checking Python 2 code as well: just pass in the --py2 flag. See Type checking Python 2 code for more details.

$ mypy --py2

Function signatures and dynamic vs static typing

A function without type annotations is considered to be dynamically typed by mypy:

def greeting(name):
    return 'Hello ' + name

By default, mypy will not type check dynamically typed functions. This means that with a few exceptions, mypy will not report any errors with regular unannotated Python.

This is the case even if you misuse the function: for example, mypy would currently not report any errors if you tried running greeting(3) or greeting(b"Alice") even though those function calls would result in errors at runtime.

You can teach mypy to detect these kinds of bugs by adding type annotations (also known as type hints). For example, you can teach mypy that greeting both accepts and returns a string like so:

def greeting(name: str) -> str:
    return 'Hello ' + name

This function is now statically typed: mypy can use the provided type hints to detect incorrect usages of the greeting function. For example, it will reject the following calls since the arguments have invalid types:

def greeting(name: str) -> str:
    return 'Hello ' + name

greeting(3)         # Argument 1 to "greeting" has incompatible type "int"; expected "str"
greeting(b'Alice')  # Argument 1 to "greeting" has incompatible type "bytes"; expected "str"

Note that this is all still valid Python 3 code! The function annotation syntax shown above was added to Python as a part of Python 3.0.

If you are trying to type check Python 2 code, you can add type hints using a comment-based syntax instead of the Python 3 annotation syntax. See our section on typing Python 2 code for more details.

Being able to pick whether you want a function to be dynamically or statically typed can be very helpful. For example, if you are migrating an existing Python codebase to use static types, it’s usually easier to migrate by incrementally adding type hints to your code rather than adding them all at once. Similarly, when you are prototyping a new feature, it may be convenient to initially implement the code using dynamic typing and only add type hints later once the code is more stable.

Once you are finished migrating or prototyping your code, you can make mypy warn you if you add a dynamic function by mistake by using the --disallow-untyped-defs flag. See The mypy command line for more information on configuring mypy.


The earlier stages of analysis performed by mypy may report errors even for dynamically typed functions. However, you should not rely on this, as this may change in the future.

More function signatures

Here are a few more examples of adding type hints to function signatures.

If a function does not explicitly return a value, give it a return type of None. Using a None result in a statically typed context results in a type check error:

def p() -> None:

a = p()  # Error: "p" does not return a value

Make sure to remember to include None: if you don’t, the function will be dynamically typed. For example:

def f():
    1 + 'x'  # No static type error (dynamically typed)

def g() -> None:
    1 + 'x'  # Type check error (statically typed)

Arguments with default values can be annotated like so:

def greeting(name: str, excited: bool = False) -> str:
    message = 'Hello, {}'.format(name)
    if excited:
        message += '!!!'
    return message

*args and **kwargs arguments can be annotated like so:

def stars(*args: int, **kwargs: float) -> None:
    # 'args' has type 'tuple[int, ...]' (a tuple of ints)
    # 'kwargs' has type 'dict[str, float]' (a dict of strs to floats)
    for arg in args:
    for key, value in kwargs:
        print(key, value)

Additional types, and the typing module

So far, we’ve added type hints that use only basic concrete types like str and float. What if we want to express more complex types, such as “a list of strings” or “an iterable of ints”?

For example, to indicate that some function can accept a list of strings, use the list[str] type (Python 3.9 and later):

def greet_all(names: list[str]) -> None:
    for name in names:
        print('Hello ' + name)

names = ["Alice", "Bob", "Charlie"]
ages = [10, 20, 30]

greet_all(names)   # Ok!
greet_all(ages)    # Error due to incompatible types

The list type is an example of something called a generic type: it can accept one or more type parameters. In this case, we parameterized list by writing list[str]. This lets mypy know that greet_all accepts specifically lists containing strings, and not lists containing ints or any other type.

In Python 3.8 and earlier, you can instead import the List type from the typing module:

from typing import List  # Python 3.8 and earlier

def greet_all(names: List[str]) -> None:
    for name in names:
        print('Hello ' + name)


You can find many of these more complex static types in the typing module.

In the above examples, the type signature is perhaps a little too rigid. After all, there’s no reason why this function must accept specifically a list – it would run just fine if you were to pass in a tuple, a set, or any other custom iterable.

You can express this idea using the (or typing.Iterable in Python 3.8 and earlier) type instead of list :

from import Iterable  # or "from typing import Iterable"

def greet_all(names: Iterable[str]) -> None:
    for name in names:
        print('Hello ' + name)

As another example, suppose you want to write a function that can accept either ints or strings, but no other types. You can express this using the Union type:

from typing import Union

def normalize_id(user_id: Union[int, str]) -> str:
    if isinstance(user_id, int):
        return 'user-{}'.format(100000 + user_id)
        return user_id

Similarly, suppose that you want the function to accept only strings or None. You can again use Union and use Union[str, None] – or alternatively, use the type Optional[str]. These two types are identical and interchangeable: Optional[str] is just a shorthand or alias for Union[str, None]. It exists mostly as a convenience to help function signatures look a little cleaner:

from typing import Optional

def greeting(name: Optional[str] = None) -> str:
    # Optional[str] means the same thing as Union[str, None]
    if name is None:
        name = 'stranger'
    return 'Hello, ' + name

The typing module contains many other useful types. You can find a quick overview by looking through the mypy cheatsheets and a more detailed overview (including information on how to make your own generic types or your own type aliases) by looking through the type system reference.


When adding types, the convention is to import types using the form from typing import Union (as opposed to doing just import typing or import typing as t or from typing import *).

For brevity, we often omit imports from typing or in code examples, but mypy will give an error if you use types such as Iterable without first importing them.


In some examples we use capitalized variants of types, such as List, and sometimes we use plain list. They are equivalent, but the prior variant is needed if you are using Python 3.8 or earlier.

Local type inference

Once you have added type hints to a function (i.e. made it statically typed), mypy will automatically type check that function’s body. While doing so, mypy will try and infer as many details as possible.

We saw an example of this in the normalize_id function above – mypy understands basic isinstance checks and so can infer that the user_id variable was of type int in the if-branch and of type str in the else-branch. Similarly, mypy was able to understand that name could not possibly be None in the greeting function above, based both on the name is None check and the variable assignment in that if statement.

As another example, consider the following function. Mypy can type check this function without a problem: it will use the available context and deduce that output must be of type list[float] and that num must be of type float:

def nums_below(numbers: Iterable[float], limit: float) -> list[float]:
    output = []
    for num in numbers:
        if num < limit:
    return output

Mypy will warn you if it is unable to determine the type of some variable – for example, when assigning an empty dictionary to some global value:

my_global_dict = {}  # Error: Need type annotation for "my_global_dict"

You can teach mypy what type my_global_dict is meant to have by giving it a type hint. For example, if you knew this variable is supposed to be a dict of ints to floats, you could annotate it using either variable annotations (introduced in Python 3.6 by PEP 526) or using a comment-based syntax like so:

# If you're using Python 3.9+
my_global_dict: dict[int, float] = {}

# If you're using Python 3.6+
my_global_dict: Dict[int, float] = {}

# If you want compatibility with even older versions of Python
my_global_dict = {}  # type: Dict[int, float]

Library stubs and typeshed

Mypy uses library stubs to type check code interacting with library modules, including the Python standard library. A library stub defines a skeleton of the public interface of the library, including classes, variables and functions, and their types. Mypy ships with stubs for the standard library from the typeshed project, which contains library stubs for the Python builtins, the standard library, and selected third-party packages.

For example, consider this code:

x = chr(4)

Without a library stub, mypy would have no way of inferring the type of x and checking that the argument to chr() has a valid type.

Mypy complains if it can’t find a stub (or a real module) for a library module that you import. Some modules ship with stubs or inline annotations that mypy can automatically find, or you can install additional stubs using pip (see Missing imports and Using installed packages for the details). For example, you can install the stubs for the requests package like this:

python3 -m pip install types-requests

The stubs are usually packaged in a distribution named types-<distribution>. Note that the distribution name may be different from the name of the package that you import. For example, types-PyYAML contains stubs for the yaml package. Mypy can often suggest the name of the stub distribution: error: Library stubs not installed for "yaml" (or incompatible with Python 3.8) note: Hint: "python3 -m pip install types-PyYAML"


Starting in mypy 0.900, most third-party package stubs must be installed explicitly. This decouples mypy and stub versioning, allowing stubs to updated without updating mypy. This also allows stubs not originally included with mypy to be installed. Earlier mypy versions included a fixed set of stubs for third-party packages.

You can also create stubs easily. We discuss ways of silencing complaints about missing stubs in Missing imports.

Configuring mypy

Mypy supports many command line options that you can use to tweak how mypy behaves: see The mypy command line for more details.

For example, suppose you want to make sure all functions within your codebase are using static typing and make mypy report an error if you add a dynamically-typed function by mistake. You can make mypy do this by running mypy with the --disallow-untyped-defs flag.

Another potentially useful flag is --strict, which enables many (though not all) of the available strictness options – including --disallow-untyped-defs.

This flag is mostly useful if you’re starting a new project from scratch and want to maintain a high degree of type safety from day one. However, this flag will probably be too aggressive if you either plan on using many untyped third party libraries or are trying to add static types to a large, existing codebase. See Using mypy with an existing codebase for more suggestions on how to handle the latter case.

Next steps

If you are in a hurry and don’t want to read lots of documentation before getting started, here are some pointers to quick learning resources:

You can also continue reading this document and skip sections that aren’t relevant for you. You don’t need to read sections in order.