Depending on your setup and preferences, there are multiple ways to run Emmett applications. In this chapter, we'll try to document the most common ones.

If you want to use an ASGI server not listed in this section, please refer to its documentation, remembering that your Emmett application object is the actual ASGI application (following spec version 3.0).

Included server

Changed in version 2.5

Emmett comes with Granian as its HTTP server. In order to run your application in production you can just use the included serve command:

emmett serve --host --port 80

You can inspect all the available options of the serve command using the --help option. Here is the full list:

option default description
host Bind address
port 8000 Bind port
workers 1 Number of worker processes
threads 1 Number of threads
threading-mode workers Threading implementation (possible values: runtime,workers)
interface rsgi Server interface (possible values: rsgi,asgi)
http auto HTTP protocol version (possible values: auto,1,2)
ws/no-ws ws Enable/disable websockets support
loop auto Loop implementation (possible values: auto,asyncio,uvloop)
log-level info Logging level (possible values: debug,info,warning,error,critical)
backlog 2048 Maximum connection queue
ssl-certfile Path to SSL certificate file
ssl-keyfile Path to SSL key file


Changed in version 2.5

In case you want to stick with a more popular option, Emmett also comes with included support for Uvicorn.

You can just use the emmett[uvicorn] extra during installation and rely on the uvicorn command to serve your application.


The included server might suit most of the common demands, but whenever you need additional features, you can use Gunicorn.

Emmett includes a Gunicorn worker class allowing you to run ASGI applications with the Emmett's environment, while also giving you Gunicorn's fully-featured process management:

gunicorn myapp:app -w 4 -k emmett.asgi.workers.EmmettWorker

This allows you to increase or decrease the number of worker processes on the fly, restart worker processes gracefully, or perform server upgrades without downtime.


Even if Docker is not properly a deployment option, we think giving an example of a Dockerfile for an Emmett application is proficient for deployment solutions using container orchestrators, such as Kubernetes or Mesos.

In order to keep the image lighter, we suggest to use the slim Python image as a source:

FROM python:3.9-slim

RUN mkdir -p /usr/src/deps
COPY requirements.txt /usr/src/deps

WORKDIR /usr/src/deps
RUN pip install --no-cache-dir -r /usr/src/deps/requirements.txt

COPY ./ /app


CMD [ "emmett", "serve" ]