Chat AI
Administrators can configure the Chat AI feature, allowing users to ask questions and chat through the AI assistant icon on the right side of the page.
Enable Chat AI
Two configuration methods are currently supported:
API: Implemented by calling a general model.Embed: Implemented by directly embedding an iframe chat window.
Using GPT model
In the right area of the top navigation bar, click .
In the left menu, click Features.
On the right page, click Chat AI tab.
In the Chat AI field, check the box to enable Chat AI.
In the Method field, select "API".
In the Types field, select "GPT".
In the Base URL field, type the API base URL of the GPT model.
In the API key field, type the API key of the GPT model.
In the Proxy field, type the proxy address if a proxy is needed to access the GPT model API.
In the GPT model field, select the GPT model to be used. Available models:
- gpt-4o-mini
- gpt-4o
- o3-mini
- o1-mini
- o1
Click Submit.
Using DeepSeek model
In the right area of the top navigation bar, click .
In the left menu, click Features.
On the right page, click Chat AI tab.
In the Chat AI field, check the box to enable Chat AI.
In the Method field, select "API".
In the Types field, select "DeepSeek".
In the Base URL field, type the API base URL of the DeepSeek model.
In the API key field, type the API key of the DeepSeek model.
In the Proxy field, type the proxy address if a proxy is needed to access the DeepSeek model API.
In the DeepSeek model field, select the DeepSeek model to be used. Available models:
- DeepSeek-V3
- DeepSeek-R1
Click Submit.
Using embed method
In the right area of the top navigation bar, click .
In the left menu, click Features.
On the right page, click Chat AI tab.
In the Chat AI field, check the box to enable Chat AI.
In the Method field, select "Embed".
In the Base URL field, type the iframe URL of the chat window.
Click Submit.