Create and distribute AI providers

This page explains the procedures for administrators to deploy and manage Agentiqs in their organization.

---

Create an AI provider

This section explains how to create an AI provider.
First, obtain the API key for the LLM you want to use.

Get the API key

The way you obtain the API key, endpoint URL, and model ID differs by AI provider.
The table below shows examples for common providers. Always refer to the official documentation for the latest information and details.

If you use Entra ID authentication, first create the Azure app registration by following Set up an Azure app for Entra ID authentication.

Model	How to obtain the API key / endpoint	Model ID (reference)
Azure OpenAI	Open the Azure OpenAI resource in the Azure portal and get the API key and endpoint URL from the "Keys and Endpoint" screen.	■
OpenAI	Sign in to the OpenAI platform and get the API key from the "API keys" page.	■
Google Gemini	Sign in to Google AI Studio and issue a project API key.	■
Microsoft Foundry	Create a project in the Azure portal or the AI Foundry portal, then create an API key from the "Keys and Endpoint" section on the overview page.	■

Configure Agentiqs

Configure Agentiqs with the obtained API key information.

Procedure

1. Click [Settings] > [AI Providers] > [Add provider].
   • The [Add AI Provider] dialog appears.
2. Select the AI provider you want to use.
   • The selected provider is added to the list.
3. Click [Edit] for the provider in the list.
   • The [Edit AI Provider] dialog appears. Configure it using the steps in Set the AI provider.
4. Select [API settings] and enter the required items.
5. Run [Test connection] in the Test connection step to verify the connection.
   • If the connection succeeds, a notification is shown.
     (If the connection fails, see Proxy settings.)
6. Click [Update] to save.

Monthly cost limits (per user)

• Monthly cost limits are determined based on the estimated usage fee. Check the official information from the AI provider you are using for actual billing amounts and the latest prices.
• For how to check estimated usage fees, see User Guide > Check AI model usage cost.

Set the AI provider

This section explains how to configure the AI provider you added.

1. API settings
   Select the [Authentication method].

   • For API-key-based authentication: Enter the API key and endpoint you obtained in Get the API key.

     
     For Azure OpenAI, enter the API version.

     The following versions are supported by Agentiqs (as of June 2026):
     2024-06-01, 2024-10-21,
     2024-08-01-PREVIEW, 2024-09-01-PREVIEW, 2024-10-01-PREVIEW, 2024-12-01-PREVIEW,
     2025-01-01-PREVIEW, 2025-03-01-PREVIEW, 2025-04-01-PREVIEW

   • For Entra ID authentication: Enter the endpoint obtained by following Set up an Azure app for Entra ID authentication.

2. Model settings and usage limits
   On the AI provider settings screen (① in the figure below), click [Edit] (②) in the edit dialog for the target AI provider.

   
   Procedure

   • From [Model settings] > [Add model], add the latest AI models.
   • From [AI provider details] > [Usage limits], set the following limits:
     • Expiration date
     • Monthly cost (per user)
   If model pricing is not set, it will not be reflected in AI model usage cost.

   Open the model settings dialog from the red-framed area below and set "Input price" and "Output price" on the [Pricing] tab.

   →

Test connection

After all settings are complete, use Test connection to confirm that the connection works.

• When you run [Test connection] on the [Model settings] tab, Agentiqs shows whether each added model can connect.
  • Confirm that the model you plan to use succeeds.
  • If you do not plan to use a model, delete it to keep the list tidy and reduce operational mistakes.
• If the models shown or the connection results differ from what you expect, review the entered settings (API key / endpoint URL / model ID).

Update model information in bulk

The model definitions for an AI provider can be bulk updated with the latest model information.
You do not need to add or edit new models one by one.

Procedure

1. In [Settings] > [AI Providers], click the edit icon for the target provider.
2. On the [Model settings] tab, click [Update to latest models] in the upper-right corner.
   
3. In the diff confirmation dialog, review the models that will be added or updated.
   To delete models that are no longer needed, check those models.
   
4. Choose "Update with this content", then choose "Update" on the detail screen to save.

Notes

• Existing display names, deployment names, and enabled/disabled states are preserved.
• Custom models not in the list are not deleted unless you check them.

Distribute AI provider settings

By distributing the connection settings and usage limits for the AI providers created by administrators,
users can import the organization-defined settings and use AI models without entering connection information individually.

There are two distribution methods. Choose the one that matches your operating model.

Distribution method	Best for	How updates are reflected	What the user does
Export file	One-time distribution, small groups	Redistribute the updated file	Import each time it is shared
Organization shared settings	Ongoing use within an organization	Automatically sync after the administrator re-exports	Register the shared settings reference once

Distribute with an export file

Export the created AI provider settings to a file and distribute it to users.

Procedure

1. Click [Settings] > [Sharing] > [Share settings] > [Export].
   • The [Export settings] dialog appears.
2. Enable the [System] check box.
   • The [AI provider] and [Proxy] check boxes are enabled automatically.
3. Enter any passkey and click [Generate export data].
   • The passkey and export data are generated and shown on screen.
4. Click [Download] to save the export data as a file.
5. Distribute the saved export data file and passkey to users.
6. Ask users to import them by following User Guide > Setting up an AI Provider.

About distribution to users

• For security reasons, we recommend distributing the passkey and export data by different means.
• We recommend enabling [Make AI provider read-only] in the [Export settings] dialog.
  When this is enabled, users in the import destination cannot change settings such as usage limits, so the settings distributed by the administrator are preserved.

Automatically distribute with organization shared settings

Settings are automatically distributed continuously through a shared folder.
Changes such as model additions, pricing updates, usage limits, and proxy settings can be reflected without user-side operations.

Procedure (for the distributing administrator)

1. Open [Settings] > [Sharing] > [Organization shared settings], then click [Export to shared folder].
2. Set the shared folder path, display name, passkey, AI providers to distribute, and other items, then export.
3. When you need to update the settings, export again to the same shared folder.

Tell users to follow User Guide > Use organization shared settings.

Notes for automatic distribution

• When re-exporting, use the same passkey as before.
• The update timing on the user side is as follows:
  • When the application starts
  • Every 10 minutes after application startup

• AI providers and models created by users themselves are not subject to automatic updates or automatic deletion.
• If proxy settings are included in the shared settings, the shared organizational settings are applied each time synchronization occurs.
  User-side proxy-specific settings are overwritten.