QClaw | MaiToken

Suitable for users who want to add an OpenAI-compatible model to QClaw and use it for code-related questions and Agent operations.

Open Model Settings: Start QClaw, open “Settings,” and find “Models,” “API,” or “Model Providers.”
Add a Provider: Click “Add,” and select OpenAI or OpenAI Compatible as the provider or protocol.
Enter the Base URL: Enter https://api.maitoken.com/v1 . QClaw will automatically append /chat/completions to this address.
Enter the API Key: Paste your MaiToken Key. Do not manually add Bearer before the Key.
Enter the Model Name: Enter <Model ID> in the Model ID field. You can set the display name to MaiToken - Model ID.
Save and Select the Model: Save the configuration, return to the chat window, and select the newly added model from the model list.

Test a Basic Conversation: Enter “Reply with OK only” in the chat box. If you receive a valid response, the API Key, Base URL, and Model ID are configured correctly.
Test Project Reading: Open a test project and enter “Read the current project directory and list only the files at the first level. Do not modify any files.”
Test Agent Capabilities: Ask the Agent to create a file named test-maitoken.txt and write hello into it. Confirm that the file was created successfully. Do not use a real production project for testing.

Issue	Solution
404 or path error	Confirm that the Base URL ends at `/v1`. Do not enter the full `/chat/completions` path.
Model list is empty	Add the exact Model ID manually, or check the current API Key's model restriction list in the MaiToken console.
Agent execution is interrupted	Confirm that the model supports streaming output and tool calling. If necessary, switch to a model suitable for code or Agent tasks.