![uvicorn](\"https://raw.githubusercontent.com/tomchristie/uvicorn/master/docs/uvicorn.png\")
\n\n \n
\nAn ASGI web server, for Python.\n
\n\n---\n\n[](https://github.com/encode/uvicorn/actions)\n[](https://pypi.python.org/pypi/uvicorn)\n\n**Documentation**: [https://www.uvicorn.org](https://www.uvicorn.org)\n\n**Requirements**: Python 3.7+ (For Python 3.6 support, install version 0.16.0.)\n\nUvicorn is an ASGI web server implementation for Python.\n\nUntil recently Python has lacked a minimal low-level server/application interface for\nasync frameworks. The [ASGI specification][asgi] fills this gap, and means we're now able to\nstart building a common set of tooling usable across all async frameworks.\n\nUvicorn supports HTTP/1.1 and WebSockets.\n\n## Quickstart\n\nInstall using `pip`:\n\n```shell\n$ pip install uvicorn\n```\n\nThis will install uvicorn with minimal (pure Python) dependencies.\n\n```shell\n$ pip install uvicorn[standard]\n```\n\nThis will install uvicorn with \"Cython-based\" dependencies (where possible) and other \"optional extras\".\n\nIn this context, \"Cython-based\" means the following:\n\n- the event loop `uvloop` will be installed and used if possible.\n- the http protocol will be handled by `httptools` if possible.\n\nMoreover, \"optional extras\" means that:\n\n- the websocket protocol will be handled by `websockets` (should you want to use `wsproto` you'd need to install it manually) if possible.\n- the `--reload` flag in development mode will use `watchgod`.\n- windows users will have `colorama` installed for the colored logs.\n- `python-dotenv` will be installed should you want to use the `--env-file` option.\n- `PyYAML` will be installed to allow you to provide a `.yaml` file to `--log-config`, if desired.\n\nCreate an application, in `example.py`:\n\n```python\nasync def app(scope, receive, send):\n assert scope['type'] == 'http'\n\n await send({\n 'type': 'http.response.start',\n 'status': 200,\n 'headers': [\n [b'content-type', b'text/plain'],\n ],\n })\n await send({\n 'type': 'http.response.body',\n 'body': b'Hello, world!',\n })\n```\n\nRun the server:\n\n```shell\n$ uvicorn example:app\n```\n\n---\n\n## Why ASGI?\n\nMost well established Python Web frameworks started out as WSGI-based frameworks.\n\nWSGI applications are a single, synchronous callable that takes a request and returns a response.\nThis doesn\u2019t allow for long-lived connections, like you get with long-poll HTTP or WebSocket connections,\nwhich WSGI doesn't support well.\n\nHaving an async concurrency model also allows for options such as lightweight background tasks,\nand can be less of a limiting factor for endpoints that have long periods being blocked on network\nI/O such as dealing with slow HTTP requests.\n\n---\n\n## Alternative ASGI servers\n\nA strength of the ASGI protocol is that it decouples the server implementation\nfrom the application framework. This allows for an ecosystem of interoperating\nwebservers and application frameworks.\n\n### Daphne\n\nThe first ASGI server implementation, originally developed to power Django Channels, is [the Daphne webserver][daphne].\n\nIt is run widely in production, and supports HTTP/1.1, HTTP/2, and WebSockets.\n\nAny of the example applications given here can equally well be run using `daphne` instead.\n\n```\n$ pip install daphne\n$ daphne app:App\n```\n\n### Hypercorn\n\n[Hypercorn][hypercorn] was initially part of the Quart web framework, before\nbeing separated out into a standalone ASGI server.\n\nHypercorn supports HTTP/1.1, HTTP/2, and WebSockets.\n\nIt also supports [the excellent `trio` async framework][trio], as an alternative to `asyncio`.\n\n```\n$ pip install hypercorn\n$ hypercorn app:App\n```\n\n### Mangum\n\n[Mangum][mangum] is an adapter for using ASGI applications with AWS Lambda & API Gateway.\n\n---\n\nUvicorn is BSD licensed code.
Designed & crafted with care.
— \ud83e\udd84 —