Skip to main content

Using Flowise

What is Flowise?

Flowise is an open-source visual tool for building customized LLM applications with a drag-and-drop interface. It allows you to create AI workflows, chatbots, and agents without writing code.

Prerequisites

Before integrating Azerion Intelligence with Flowise:

Integration Steps

1. Configure ChatOpenAI Node

Since Flowise supports OpenAI-compatible APIs, you can use Azerion Intelligence through the ChatOpenAI node:

  1. Add ChatOpenAI Node:

    • Drag a "ChatOpenAI" node from the Chat Models section
    • This node supports OpenAI-compatible APIs
  2. Configure the Node:

    {
    "modelName": "meta.llama3-3-70b-instruct-v1:0",
    "temperature": 0.7,
    "maxTokens": 1024,
    "openAIApiKey": "YOUR_AZERION_API_KEY",
    "basePath": "https://api.azerion.ai/v1"
    }

2. Set Up Credentials

  1. Create Credential:

    • Go to Settings → Credentials
    • Click "Add New Credential"
    • Select "OpenAI API" credential type
  2. Configure Credential:

    {
    "credentialName": "Azerion Intelligence",
    "openAIApiKey": "YOUR_AZERION_API_KEY",
    "basePath": "https://api.azerion.ai/v1"
    }
  3. Apply to Node:

    • Select the credential in your ChatOpenAI node

Basic Example

Create a simple chatbot:

Flow Structure:

Simple Chatbot Flow
├── ChatOpenAI (Azerion Intelligence)
├── Buffer Window Memory (k=5)
└── Input/Output

Configuration:

  • ChatOpenAI Node: Use meta.llama3-3-70b-instruct-v1:0 model
  • Memory Node: Buffer Window Memory with k=5 for conversation context
  • System Message: "You are a helpful AI assistant"

Troubleshooting

API Connection Issues

Problem: Failed to connect to model Solution:

  1. Verify your API key is correct
  2. Check the base URL is set to https://api.azerion.ai/v1
  3. Ensure the model name is spelled correctly: meta.llama3-3-70b-instruct-v1:0

Authentication Errors

Problem: 401 Unauthorized errors Solution:

  1. Regenerate your API key from https://app.azerion.ai/account#api-tokens
  2. Update the credential in Flowise with the new key
  3. Make sure the API key is stored securely in Flowise's credential manager

Model Response Issues

Problem: Empty or error responses Solution:

  1. Check your prompt format and ensure it's not empty
  2. Verify the temperature and max token settings are reasonable
  3. Test with a simple prompt first to confirm the connection works