Show HN: Padwan-LLM, a lightweight LLM Python client

Category: ai-ml

Tags: llm-client, python, async, mcp, opentelemetry

Score: 6.5/10 (Innovation: 5, Technical: 7, Documentation: 8, Utility: 6)

Padwan-LLM is a lightweight async Python client that unifies access to OpenAI, Gemini, Mistral, Grok, Anthropic, and any OpenAI-compatible endpoint. It stands out with built-in MCP transports, realtime speech-to-speech sessions, agentic loops with tool dispatch, and opt-in OpenTelemetry/Langfuse instrumentation, all on a single HTTP dependency (niquests) with automatic HTTP/2 and HTTP/3 negotiation.

Target audience: Python backend and AI application developers integrating multiple LLM providers

Repository: https://github.com/polarsen-io/padwan-llm · Python · MIT · 3 stars

View on Hacker News