Simple hello world request is actually 🧂 3x faster than Fast API (RPS wise) on my machine 🫣.
20k rps vs 60k+ rps.
Both using 4 workers.
Fast API with uvicorn 4 workers.
Django bolt with 4 python workers and 1 actix worker.
Too avoid GIL we use 4 python processes not threads to handle requests our actix worker.
From: @Farhan Ali Raza
Mastodon
Farhan Ali Raza (@farhanaliraza@mastodon.social)
My Django API at the speed of light effort
https://github.com/FarhanAliRaza/django-bolt
It got hindered by stupid StreaminResponse
For every gen...