OMNILOGIN DOCS

English

Tiếng Việt

English

Tiếng Việt

Appearance

Custom AI ​

This node supports generating text based on input. Users can customize the AI ​​as they wish. It automatically moves to the next model if the previous one is not working.

Model Name ​

The name of the model to use. Users can find the AI ​​model by accessing this Model page to search for the desired model IDs.

URL ​

This is the URL of the AI ​​you want to use. To find out the correct value of an AI, users should search using the keywords AI Name + API documentation to display the official documentation page of the AI ​​and then search within that page to find the URL.

Some URLs of popular AIs:

  • ChatGPT: https://api.openai.com/v1
  • Claude: https://api.anthropic.com/v1
  • Grok: https://api.x.ai/v1
  • DeepSeek: https://api.deepseek.com

API Key ​

The API key is used for the AI ​​the user wants to use. To find the API creation page, users can search using the following keywords: AI Name + Developer Dashboard

Prompt ​

Present the request when creating a text snippet or an image

Assign to a variable ​

You can assign a value to a variable.

  • Variable Name: The name of the variable to which to assign the value. This field appears when the user selects to assign to the variable.

Insert into a table ​

You can insert a value into a column in table

  • Select column: The column where the value is inserted. This field appears when the user selects to insert into a table.

Add additional rows ​

Insert an additional row into a column in table.

Practical Example ​

For example, to request the ChatGPT AI to write questions for the Professor in the Kite AI testnet project, use the following:

i

To do this, the user will need to configure the ChatGPT AI node as follows:

To create this API, the user accesses the API Key page.

Create a new API key:

api

Enter a name and click create.

api

Save that API address for later use.

api

Then create the Custom AI node, and then select... Pen icon in the model for editing

api

In the Model Name section, since the user wants to create a question, they will use the gpt-4o-mini model, which is suitable for creating questions. Next, fill in the URL corresponding to the ChatGPT AI. Finally, fill in the API to use that model.

api

Next, in the Prompt section, since the user wants to create a question to ask the Professor AI, which helps the user explore the Kite AI ecosystem, the user will describe the prompt as follows: Create a question related to the Kite AI testnet ecosystem. Additionally, the user can add requirements such as question in English, not exceeding 100 characters. Return only the question without adding any characters, then the user will fill in as follows:

i

Finally, after this node runs, it will return a question. To receive that question, we need to use a variable/table. For ease of use in the process, the user will assign the question to a variable and name that variable questionAI.

i

After configuration, the user will run the node, and the AI ​​will return the result in the questionAI variable as follows:

i

After obtaining the question, the user will write the question to the send field using the node's Press Key function, using the expression to retrieve the variable value {{variables.questionAI}} and send it to the AI. AI

i

Then the paragraph will be displayed as follows:

i

Thus, the ChatGPT node was used to create the question in the Kite AI testnet.