Skip to content

Async Support

HTTPX offers a standard synchronous API by default, but also gives you the option of an async client if you need it.

Async is a concurrency model that is far more efficient than multi-threading, and can provide significant performance benefits and enable the use of long-lived network connections such as WebSockets.

If you're working with an async web framework then you'll also want to use an async client for sending outgoing HTTP requests.

Making Async requests

To make asynchronous requests, you'll need an AsyncClient.

>>> async with httpx.AsyncClient() as client:
>>>     r = await client.get('https://www.example.com/')
>>> r
<Response [200 OK]>

Tip

Use IPython or Python 3.8+ with python -m asyncio to try this code interactively, as they support executing async/await expressions in the console.

API Differences

If you're using an async client then there are a few bits of API that use async methods.

Making requests

The request methods are all async, so you should use response = await client.get(...) style for all of the following:

  • AsyncClient.get(url, ...)
  • AsyncClient.options(url, ...)
  • AsyncClient.head(url, ...)
  • AsyncClient.post(url, ...)
  • AsyncClient.put(url, ...)
  • AsyncClient.patch(url, ...)
  • AsyncClient.delete(url, ...)
  • AsyncClient.request(method, url, ...)
  • AsyncClient.send(request, ...)

Opening and closing clients

Use async with httpx.AsyncClient() if you want a context-managed client...

async with httpx.AsyncClient() as client:
    ...

Alternatively, use await client.aclose() if you want to close a client explicitly:

client = httpx.AsyncClient()
...
await client.aclose()

Streaming responses

The AsyncClient.stream(method, url, ...) method is an async context block.

>>> client = httpx.AsyncClient()
>>> async with client.stream('GET', 'https://www.example.com/') as response:
>>>     async for chunk in response.aiter_bytes():
>>>         ...

The async response streaming methods are:

  • Response.aread() - For conditionally reading a response inside a stream block.
  • Response.aiter_bytes() - For streaming the response content as bytes.
  • Response.aiter_text() - For streaming the response content as text.
  • Response.aiter_lines() - For streaming the response content as lines of text.
  • Response.aiter_raw() - For streaming the raw response bytes, without applying content decoding.
  • Response.aclose() - For closing the response. You don't usually need this, since .stream block closes the response automatically on exit.

For situations when context block usage is not practical, it is possible to enter "manual mode" by sending a Request instance using client.send(..., stream=True).

Example in the context of forwarding the response to a streaming web endpoint with Starlette:

import httpx
from starlette.background import BackgroundTask
from starlette.responses import StreamingResponse

client = httpx.AsyncClient()

async def home(request):
    req = client.build_request("GET", "https://www.example.com/")
    r = await client.send(req, stream=True)
    return StreamingResponse(r.aiter_text(), background=BackgroundTask(r.aclose))

Warning

When using this "manual streaming mode", it is your duty as a developer to make sure that Response.aclose() is called eventually. Failing to do so would leave connections open, most likely resulting in resource leaks down the line.

Streaming requests

When sending a streaming request body with an AsyncClient instance, you should use an async bytes generator instead of a bytes generator:

async def upload_bytes():
    ...  # yield byte content

await client.post(url, data=upload_bytes())

Supported async environments

HTTPX supports either asyncio or trio as an async environment.

It will auto-detect which of those two to use as the backend for socket operations and concurrency primitives.

AsyncIO

AsyncIO is Python's built-in library for writing concurrent code with the async/await syntax.

import asyncio
import httpx

async def main():
    async with httpx.AsyncClient() as client:
        response = await client.get('https://www.example.com/')
        print(response)

asyncio.run(main())

Trio

Trio is an alternative async library, designed around the the principles of structured concurrency.

import httpx
import trio

async def main():
    async with httpx.AsyncClient() as client:
        response = await client.get('https://www.example.com/')
        print(response)

trio.run(main)

Important

The trio package must be installed to use the Trio backend.

Calling into Python Web Apps

Just as httpx.Client allows you to call directly into WSGI web applications, the httpx.AsyncClient class allows you to call directly into ASGI web applications.

Let's take this Starlette application as an example:

from starlette.applications import Starlette
from starlette.responses import HTMLResponse
from starlette.routing import Route


async def hello():
    return HTMLResponse("Hello World!")


app = Starlette(routes=[Route("/", hello)])

We can make requests directly against the application, like so:

>>> import httpx
>>> async with httpx.AsyncClient(app=app, base_url="http://testserver") as client:
...     r = await client.get("/")
...     assert r.status_code == 200
...     assert r.text == "Hello World!"

For some more complex cases you might need to customise the ASGI transport. This allows you to:

  • Inspect 500 error responses rather than raise exceptions by setting raise_app_exceptions=False.
  • Mount the ASGI application at a subpath by setting root_path.
  • Use a given client address for requests by setting client.

For example:

# Instantiate a client that makes ASGI requests with a client IP of "1.2.3.4",
# on port 123.
transport = httpx.ASGITransport(app=app, client=("1.2.3.4", 123))
async with httpx.AsyncClient(transport=transport, base_url="http://testserver") as client:
    ...

See the ASGI documentation for more details on the client and root_path keys.