3 ms·
Ask HN: How do you deploy Python-based LangChain or LlamaIndex code on prod?
- efriis 2y agoErick from LangChain here - would recommend either deploying with LangGraph Cloud or anywhere that deploys a FastAPI easily. LangGraph Cloud: https://langchain-ai.github.io/langgraph/cloud/ https://langchain-ai.github.io/langgraph/cloud/ Example (chat-langchain): https://github.com/langchain-ai/chat-langchain?tab=readme-ov-file https://github.com/langchain-ai/chat-langchain?tab=readme-ov... FastAPI example with LangServe (old chat-langchain implementation): https://github.com/langchain-ai/chat-langchain/blob/langserve/backend/main.py https://github.com/langchain-ai/chat-langchain/blob/langserv... Or just use regular fastapi endpoints instead of langserve.add_routes with a fastapi server: https://fastapi.tiangolo.com/ https://fastapi.tiangolo.com/