Connecting LibreChat with LangFlow via Proxy

[cortado533]

Hello everyone,

I'm currently using LibreChat as a chat frontend, which can access interfaces built upon the OpenAI API standard. I would like to connect LangFlow and LibreChat together to leverage the capabilities of both tools.

I believe that to achieve this, I might need to build something like a proxy to interface between the two applications.

• Has anyone had experience with this setup?
• How could one approach building such a proxy?
• Is there an existing solution or best practice for connecting LibreChat with LangFlow?
• Is this a valid use case, or are there better alternatives to achieve the integration I'm aiming for?

Any insights or suggestions would be greatly appreciated!

Thank you!

[dosubot[bot]]

Hey @cortado533! I'm here to help you with any bugs, questions, or contributions. Let's tackle this challenge together!

I found a similar discussion that might be helpful for your use case. In the discussion, users explored deploying Langflow as an API and interacting with it using API endpoints ^[1].

To connect LibreChat with Langflow via a proxy, you can follow these steps:

1. Deploy Langflow as an API:

   • Prepare a POST request to the /run/{flow_id} endpoint, replacing {flow_id} with the ID of the flow you wish to execute.
   • Include necessary headers in your request, such as Content-Type: application/json and x-api-key: YOUR_API_KEY for authentication.
   • Construct the request body with the input value and any tweaks.

2. Build a Proxy:

   • Create a simple proxy server using a web framework like Flask (Python) or Express (Node.js).
   • The proxy server will receive requests from LibreChat, forward them to the Langflow API, and return the response back to LibreChat.

Here is an example using Flask:

from flask import Flask, request, jsonify
import requests

app = Flask(__name__)

LANGFLOW_API_URL = "http://<your_server>/api/v1/run/{flow_id}"
API_KEY = "YOUR_API_KEY"

@app.route('/proxy', methods=['POST'])
def proxy():
    data = request.json
    headers = {
        "Content-Type": "application/json",
        "x-api-key": API_KEY
    }
    response = requests.post(LANGFLOW_API_URL, json=data, headers=headers)
    return jsonify(response.json())

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=5000)

3. Configure LibreChat:
   • Point LibreChat to your proxy server endpoint (/proxy in the example above).

This setup allows LibreChat to interact with Langflow through the proxy server, leveraging the capabilities of both tools.

For more details on deploying Langflow as an API, you can refer to the discussion ^[1].

If you encounter any specific issues or need further customization, feel free to ask!

To continue talking to Dosu, mention @dosu.

[cortado533]

@dosu Does this also work with streaming?

[dosubot[bot]]

Yes, the proxy setup for connecting LibreChat with LangFlow supports streaming. To enable streaming, you can append the ?stream=true query parameter to the request. This allows LLM chat responses to be streamed back as token events until the end event closes the connection. You can test this setup using a curl command like this:

curl -X POST \
  "$LANGFLOW_URL/api/v1/run/$FLOW_ID?stream=true" \
  -H "accept: application/json" \
  -H "Content-Type: application/json" \
  -d '{
    "message": "Tell me something interesting!",
    "session_id": "chat-123"
  }'

This setup is consistent with the streaming: bool = True configuration in the ChatConfig class ^[1][2].

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}