FlowiseAI

Step 1: Add CometAPI Node and Configure Credentials#

1.

In the FlowiseAI canvas, click Add new.

2.

In the search box, type and search for "cometapi" to find the CometAPI box (i.e., the ChatCometAPI node).

3.

Drag the CometAPI box onto the workspace canvas.

4.

In the Connect Credential dropdown menu of the CometAPI node, select Create New.

5.

Enter the API Key (sk-xxxxx) you obtained from CometAPI.

Optional Configuration: Set the Model Name (e.g., gpt-4o-mini), Temperature (e.g., 0.7, to control response randomness), and Additional Parameters (add other parameters as needed).

Step 2: Add Prompt Node and Set Up LLM Chain#

1.

In the + (Add Nodes) menu, search for "Prompt", then find and select the Prompt Template node.

2.

Drag the Prompt Template node onto the workspace canvas.

3.

In the Prompt Template node, set up the prompt template, for example:

Question: {question}
Answer: Let's think step by step.

This will create a basic prompt structure to guide the model's response.

4.

Next, set up the chain model:

In the + (Add Nodes) menu, search for "Chain", then find and select the LLM Chain node.

Drag the LLM Chain node onto the workspace canvas.

Configure the LLM Chain:

Language Model: Select or connect to the CometAPI node from Step 1.

Prompt: This will be connected to the Prompt Template node later.

Output Parser: Optional, used to parse the output format.

Chain Name: Enter a name for your chain, for example, "Universal Chat Assistant".

Input Moderation: Optional, enable input checks for safety.

Step 3: Connect Nodes and Complete the Flow#

Troubleshooting#

Summary#