302.AI launches a new feature - Custom API!
In order to make it easier for users to connect third-party APIs to the 302.AI platform, we have launched the Custom API feature. With this feature, users can easily call multiple external APIs through a single interface, achieve rapid integration with major platforms, and greatly improve development efficiency.
Feature highlights:
API transfer: quickly integrate third-party APIs into the 302.AI platform to quickly achieve multi-tenant capabilities
Generate MCP Server: The integrated third-party API can be generated into MCP Server with one click, and quickly connected to 302 chatbots or third-party clients for use
Price description: Free
The Custom API feature is open to all users at no additional cost.
302.AI is committed to providing developers with more efficient and convenient tools to help quickly build intelligent solutions. Experience the Custom API feature now and unlock more possibilities!
How to access custom API in 302.AI?
一、Create an API project
1、Manually create an API project
Log in to 302.AI and select External Resources > Custom API
Click Create API Project and follow the steps to fill in the relevant configuration information
(1) Fill in the project information, API project name (required), API project description
(2) Configure project parameters, including Request Domain (required), Translation path, Forward Location, Authentication Method (required), and Auto Forward
Note: Auto forward
On: All API paths of the request domain will be automatically forwarded, and specific paths can be disabled in the API list;
Off: Only manually added and enabled API paths are forwarded
Successfully created API project
2、Creating an API project by importing data
(1) Fill in the project information, including API project name (required), API project description, and import data.
(2) Configure project parameters, including Request Domain (required), Translation Path, Forward Location, Authentication Method (required), and Auto Forward, and click Create
Note: Auto forward
On: All API paths of the request domain will be automatically forwarded, and specific paths can be disabled in the API list;
Off: Only manually added and enabled API paths are forwarded
The data was imported successfully, the API project was created successfully, click Confirm
View the successfully imported API
二、Adding API
Click Operation > View, click Add API, and follow the steps to fill in the relevant configuration information
1、Manually add API
(1) Fill in the API information, including API name (required), Function Name (required), API description, request domain > request path (required), request method, and click Next Step.
Note: Function Name is used to call MCP Server, similar to mcptoolname. To improve LLM understanding, it is recommended that the name be consistent with the function. Only letters, numbers, and underscores are supported.
(2) Configure API parameters, configure Input Parameters (Params, Body, Header), and Response Body
Note: Body supports switching to JSON editing
(3) Test, fill in parameter values for testing
Note: This step is optional and the test can be performed or skipped
Add API successfully
2、Adding API via data import
OpenAPI URL import or document import
Note: Supports importing json or yaml files in OpenAPI 3, Swagger 1, 2, 3 data formats (click here to view data import tutorial)
Import API successfully
How to call custom API in MCP Server?
Select Agent > MCP Server, enter the Server name, and click Configure
Click Custom API Tools, select an API project, check the custom API tools you need to use, and click Confirm
Click Add Server. After adding successfully, click Integrate and CopyURL
Call the custom API tool in the chatbot (click here to see how to call the MCP Server)