Tag
#python
12 open-source projects filed under this tag.
aisuite
Andrew Ng's aisuite is a unified Python library for building with multiple LLM providers — one API for OpenAI, Anthropic, Google, Mistral, and more, plus an Agents API with toolkits and MCP support.
django-bolt
Django-Bolt is a Rust-powered API framework for Django that hits 188k+ RPS — faster than FastAPI with full Django ORM and admin support.
enrichmcp
EnrichMCP is a Python ORM for AI agents that turns data models into typed, navigable MCP tools with relationship traversal and Pydantic validation.
iii
Rust-based service composition framework with Worker/Function/Trigger primitives, SDKs for TypeScript, Python, Rust, and Go, real-time observability, and native agent support.
markitdown
A Python tool for converting files and office documents to Markdown. Supports PDF, PowerPoint, Word, Excel, images, audio, HTML, and more — all with a clean, unified output format.
microsoft-agent-governance-toolkit
Microsoft's open-source toolkit for governing autonomous AI agents — policy enforcement, zero-trust identity, execution sandboxing, and audit logging. Covers 10/10 OWASP Agentic Top 10.
nanobot
Nanobot is a lightweight open-source Python AI agent with WebUI, multi-channel chat, MCP support, and memory — 43K+ stars on GitHub.
onyx
Onyx is an open-source AI platform with RAG, agents, web search, and code execution — a self-hosted alternative to ChatGPT Enterprise built with Next.js and Python.
pydantic-monty
Pydantic Monty is a minimal Python interpreter written in Rust for safely executing LLM-generated code without containers. Fast startup, sandboxed by design.
scrapling
Adaptive Python web scraping framework with anti-bot bypass, smart element tracking, and full crawl capabilities for modern web scraping.
tokenspeed
TokenSpeed is an open-source LLM inference engine built for agentic workloads — 580 tokens/sec on Qwen3.5-397B with TensorRT-LLM performance and vLLM-level ease of use.
turbovec
Turbovec is a Rust vector index with Python bindings built on Google's TurboQuant algorithm — faster than FAISS with zero training phase and 16x compression.