Chat App Integration

The platform can push your configuration into a range of AI chat applications for testing and everyday use.

Integration methods

There are three:

• One-click import from a chat preset
• One-click import from an API key
• Manual chat app configuration

Before you start

1. Sign in to the platform and create a key on the API Keys page.
2. Confirm the key is enabled, has enough quota, has not expired, and that the current IP, group and model restrictions allow the call.

See API Keys for how to create and use keys.

One-click import from a chat preset

1. Click Chat in the left sidebar to expand the presets configured on the platform.
2. Click the preset you want.
3. The browser tries to launch the installed external application in a new window.
4. The system switches to the target application and fills in the configuration.

Note

The preset list grows as more applications are supported. If there is no Chat module in the sidebar at all, it is most likely hidden in your profile settings, see the common problems below.

One-click import from an API key

1. Open API Keys in the left sidebar.
2. Select the API key you want to use.
3. Next to the selected key, click … → Chat → the target chat application.
4. You jump straight to that application with the selected key imported.

Manual chat app configuration

Applications that do not support one-click import can be configured by hand.

Collect the following from the platform:

Setting	Description	Where to find it
API base URL	The API endpoint address	The platform address, for example https://api.example.com/v1
API key	The API key	The API Keys page
Available models	The list of supported models	Filter the Model Marketplace by the group your key belongs to

General configuration steps

1. Open the Settings or Configuration area in the target application.
2. Find the Custom API or OpenAI-compatible API section.
3. Enter the API base URL and the API key.
4. Save the configuration and start using it.

Most applications that support the OpenAI-compatible API can connect by entering a custom API address. If the application expects a full endpoint, use https://api.example.com/v1/chat/completions and add any extra path segments the application’s documentation calls for.

Common problems

• No chat presets in the sidebar: check that the Chat module is enabled under Profile → Personal sidebar settings.
• “No keys available”: create or enable an API key, and check quota, expiry, model restrictions and the IP allowlist.
• The desktop application does not launch: install the target application first and confirm the operating system has registered its custom protocol. The browser may also ask you to confirm opening an external application.
• A web preset is blank or refuses to load: the target site may block iframes via CSP or X-Frame-Options. Use that application’s manual configuration instead.
• An extension client connects but the key is invalid: confirm the extension is installed and active on the current page. If it still fails, use manual configuration or another preset.
• Requests return unauthorized or forbidden: check the auth header format, model name, key state, group permissions and IP restrictions. Claude and Gemini requests also need the matching path and headers.