DifyAI
Connect your virtual being to a DifyAI agent for LLM-powered conversational responses with tool use and knowledge retrieval.
DifyAI is an LLM application platform that lets you build, deploy, and manage AI agents with tool calling, knowledge retrieval, and custom workflows. When integrated with Virbe, your virtual being forwards user messages to a DifyAI agent and returns the agent's response.
Prerequisites
- An active DifyAI account (cloud or self-hosted)
- A published DifyAI application (Chatbot or Agent type)
- The application's API key
Configuration
Credentials
| Field | Description |
|---|---|
| API key | The API key for your DifyAI application. Found in the DifyAI dashboard under your application → API Access |
| Base URL | The DifyAI API endpoint. Use https://api.dify.ai/v1 for DifyAI Cloud, or your self-hosted instance URL |
Finding Your API Key
- Log in to your DifyAI dashboard
- Open the application you want to connect
- Go to API Access in the left sidebar
- Copy the API Secret Key
How It Works
When the Call Conv AI node in a Conversational Flow routes to the DifyAI engine:
- Virbe sends the user's message to the DifyAI
/chat-messagesendpoint - DifyAI processes the message through your configured agent (including any tools, knowledge bases, or workflows you've set up in DifyAI)
- DifyAI returns the agent's response text
- Virbe delivers the response to the user
The conversation history is managed via DifyAI's native conversation threading.
Use Cases
DifyAI is a strong choice when you want to:
- Leverage DifyAI's built-in knowledge base and RAG capabilities alongside or instead of Virbe's Knowledge Base
- Use DifyAI's workflow orchestration for complex multi-step logic
- Run a self-hosted LLM with DifyAI as the orchestration layer
- Connect to DifyAI tool integrations (web search, database queries, etc.)
Ensure your DifyAI application is in Published state before connecting it to Virbe. Draft applications will not respond to API requests.