Plans supporting the use of Ask Eddy AI API
Professional | Business | Enterprise |
---|---|---|
The Eddy AI Public API is an invaluable feature designed to enhance the functionality and flexibility of our Eddy AI search feature. The Eddy AI Public API is particularly useful for businesses looking to deeply integrate AI-driven search capabilities directly into their own platforms or services. By embedding Eddy AI into your product, you provide users with the ability to access powerful search functionality without needing to visit your documentation site.
This API offers greater flexibility in customizing the user interface, allowing you to tailor aspects like color, naming, and placement to align with your brand and user experience. With this seamless integration, Eddy AI becomes a natural extension of your product, enhancing user interactions and improving search efficiency. By exposing the API endpoints, you can incorporate them into your platforms, effortlessly extract answers from Eddy AI and build a customized search experience.
NOTE
Other than English (en-US, en-GB), the Eddy AI Public API supports five more languages such as
French (fr)
Spanish (es)
German (de)
Norwegian (no)
Dutch (nl)
Using the Public API
Navigate to the Settings () > AI Features > Eddy AI in the Knowledge base portal.
In AI assistive search, select the Public API checkbox.
NOTE
Ensure at least one of the checkboxes is selected. Otherwise, the AI assistive search toggle will be automatically disabled.
If the Public API is not enabled, you'll receive a 500 response code error while using Ask Eddy AI API.
Each API call will reduce your total credit count by one.
Create authorization API token.
NOTE
In the Knowledge Base portal, navigate to Settings () > Knowledge Base Portal > API Tokens.
Click Create API Token, enter the token name, and specify the POST method.
Click Create and copy the generated API token.
You can find the API structure at the API hub.
Try the Eddy AI Public API in our API docs
Navigate to the Document360 API documentation.
On the right side of the page, click the Try it section.
In the Token field, field, enter the API token generated from your Document360 project.
Enter the desired values in the Body section. The default request box will appear as follows:
Prompt: Enter the question or query.
Version ID: Obtain the current version ID.
Language Code: Specify the desired language code for the input.
Click Try it & see response and the response will be generated.
Try the Eddy AI Public API in Swagger
Navigate to the Swagger API hub. Under project versions, you can find the Ask Eddy AI API.
In the top right of the Swagger API hub and click Authorize.
The Available authorizations panel will appear, prompting you to enter the API token.
Paste the API token into the Available Authorizations window and click Authorize.
Once authorized, navigate to Project versions >
/v2/ProjectVersions/ask-Eddy AI
and click Try it out.The default request box will appear as follows:
Prompt: Enter the question or query.
Version ID: Obtain the current version ID from the
/v2/ProjectVersions
endpoint.Language Code: Specify the desired language code.
Click Execute to receive a successful response containing all required information.
NOTE
If you receive a service unavailable message, ensure the Public API is enabled in the settings.
FAQ
Why is there a discrepancy in the source articles between the Ask Eddy public API and Eddy AI search on the user site?
The discrepancy in source articles may occur if you're retrieving details using the articleID
for an article that is of type "Pagecategory." In such cases, fields like article_id
, article_title
, and article_slug
will return null in the API response.
For "Pagecategory" articles, the details should be fetched using the categoryID
instead of the articleID
. Here's an example of what the API response would look like for a "Pagecategory" article:
{
"category_id": "_categoryid",
"category_title": "Title",
"category_slug": "slug",
"version_name": "v1",
"version_slug": "v1",
"article_id": null,
"article_title": null,
"article_slug": null,
"version_display_name": null
}
To resolve this issue, ensure that you're using categoryID
to retrieve information for "Pagecategory" articles when using the Eddy API.