chenki

Self-hostable LLM client for restaurant tech. A thin Python library that talks to any OpenAI-compatible /v1/chat/completions endpoint, with restaurant-domain helpers (menu Q&A, dish classification, order parsing) built on top.

By default it points at brianchenhao-chenki-llm.hf.space — a free Hugging Face Space running Qwen 2.5 1.5B Instruct (Q4_K_M) behind llama.cpp. Self-hosters can deploy their own Space using the included server/Dockerfile.

Status

chenki is under active development. v0.0.1 reserves the PyPI name; v0.1.0 is the first functional release.

Install

pip install chenki

Quickstart

from chenki import ChenkiClient, Message

client = ChenkiClient()  # defaults to https://brianchenhao-chenki-llm.hf.space/v1
reply = client.chat([Message(role="user", content="Hello!")])
print(reply.text)

Website

chenki.com (coming soon)

License

MIT