awesome-python
An opinionated list of Python frameworks, libraries, tools, and resources
Awesome Python: a hub of curated Python resources
Awesome Python is an opinionated directory of Python frameworks, libraries, and tools, organized by category. The README notes it ranks among the ten most-starred repositories on GitHub and sells sponsorship space.
A directory with sponsors
The repository bills itself as an opinionated list of Python frameworks, libraries, tools, and resources. The README opens with a sponsorship offer, noting the project sits among the ten most-starred repositories on GitHub and that sponsors get their product in front of Python developers. That is the entire business angle. The rest of the file is a long, categorized index of projects.
AI, deep learning, and machine learning
The AI and agents section splits into agent skills and orchestration. On the skills side sit django-ai-plugins for Django code review and sentry-skills for Python-focused engineering. Orchestration covers ag2, an open source AgentOS, autogen for building agentic applications, and bernstein, a deterministic orchestrator for CLI coding agents. The deep learning section names jax, keras, pytorch, pytorch-lightning, stable-baselines3, and tensorflow. Machine learning brings in catboost, lightgbm, xgboost, scikit-learn, and spark.ml, among others.
Web frameworks and the API layer
Web frameworks are split into synchronous and asynchronous. Bottle, django, flask, and pyramid sit in the first group; litestar, reflex, robyn, starlette, and tornado in the second. API libraries get their own breakdown by Django, Flask, and framework-agnostic. That last group includes connexion, falcon, fastapi, sanic, and strawberry. Web servers are separated into ASGI and WSGI, with uvicorn and gunicorn as the familiar names.
The long tail of categories
The list does not stop at the web. It covers WebSocket libraries, template engines like jinja, authentication tools such as authlib and django-allauth, admin panels, CMS systems, and HTTP clients including requests and httpx. NLP gets gensim, nltk, and spacy. Computer vision gets opencv and ultralytics. Everything is a one-line entry with a link, and nothing in this repository is code. It is a catalog.
Community notes