Hi! I'm looking for a demo code example that demonstrates how to connect VAPI with LangChain. Could you provide a simple implementation or guide me on how to get started? Thank you!

Hi! There is an option to use your own server for answering the calls https://docs.vapi.ai/customization/custom-llm/using-your-server And here is the snippet in Python that I've found to get started https://replit.com/@jannismoore/Vapi-Custom-LLM?v=1#main.py Though it has some bugs, and looks like Vapi has updated their incoming payload a little bit

hi i checked it through this route, it dosen't work, it dose work through tools but there is a bit of lag, i was hoping if i could find a better way, thanks.

I need more information to better understand: 1.What is your specific use case for integrating VAPI with LangChain? \- Are you trying to use LangChain for specific capabilities like agents, chains, or memory? \- What kind of functionality are you trying to achieve? 2\. When you say there's "lag" with the tools approach: \- How much latency are you experiencing? \- At what point in the interaction does the lag occur? 3\. For the custom LLM server approach that didn't work: \- What specific issues did you encounter? \- Were there any error messages?

it's more specifically to connect langgraph to have more low level control of the agent. i already use langgraph and am able to connect it through tools but i wanted to see if there is a better way to connect as i am not looking to reinvent the wheel.

Hello Mediumreach, could you please elaborate on the reasons why you were unable to encapsulate your solution within a custom LLM and subsequently utilize it?

i tried to connect langchain/langgraph with custom llm, there is absolutely nothing available online as a refrence which works and i tried everything myself, it connects to different models that way but it dosen't connect to langchain and langgraph that way.