Differences and uses between WSGI, CGI, FastCGI, and mod_python in regards to Python?

A part answer to your question, including scgi.

• What's the difference between scgi and wsgi?
• Is there a speed difference between WSGI and FCGI?
• How Python web frameworks, WSGI and CGI fit together

CGI vs FCGI

Lazy and not writing it on my own. From the wikipedia: http://en.wikipedia.org/wiki/FastCGI

Instead of creating a new process for each request, FastCGI uses persistent processes to handle such requests. Multiple processes can configured, increasing stability and scalability. Each individual FastCGI process can handle many requests over its lifetime, thereby avoiding the overhead of per-request process creation and termination

There's also a good background reader on CGI, WSGI and other options, in the form of an official python HOWTO: http://docs.python.org/2/howto/webservers.html

In a project like Django, you can use a WSGI (Web Server Gateway Interface) server from the Flup module.

A WSGI server wraps a back-end process using one or more protocols:

• FastCGI (calling a server process)
• SCGI (Simple Common Gateway Interface - a simpler FastCGI)
• AJP (Apache JServ Protocol - a Java FastCGI)
• mod_python (runs pre-loaded code per request - uses lots of RAM)
• CGI (Common Gateway Interface, starts a process per request - slow)

In 2019, WSGI was superseded by ASGI (Asynchronous Server Gateway Interface), used by frameworks like FastAPI on servers like Uvicorn, which is much faster.

• FastCGI is a kind of CGI which is long-live, which will always be running.
• With FastCGI, it'll take less time.
• Because of multi-processes, FastCGI will cost more memory than CGI.

In Detail Diff between FastCGI vs CGI