Claude Desktop App#
Connect to the AICodeWith service by configuring the Gateway URL, API key, and Model ID.Claude Desktop App Configuration#
macOS / Linux#
macOS or Linux requires Git to be installed first.
Install Git#
On macOS, you can choose the Homebrew installation method. Some resources required during Homebrew installation are hosted overseas, so you may need to enable a TUN proxy.Click to copy the command and run it in Terminal.Run the installation command in Terminal:1. Download the Claude App#
The download page requires access to the international internet.
2. Enable Developer Mode#
After installation, open the Claude App. In the top-left corner of the screen, click Help -> Troubleshooting -> Enable Developer Mode, then select Enable in the pop-up.3. Open Third-Party Inference Configuration#
The app will restart automatically. After it restarts, click Developer -> Configure Third-Party Inference in the top-left corner of the screen to open the configuration page.4. Fill in the Gateway Base URL and Gateway API Key#
On the Connection page, enter the Gateway base URL and Gateway API key, then set Gateway auth scheme to x-api-key.5. Add Models#
Scroll down to the Model list section, click Add to add a model, and enter a Model ID, for example:claude-haiku-4-5-20251001
The display name is the name shown in the app. You can use a short name for the corresponding model.Finally, click Apply Changes to apply the changes. When prompted to restart, select Save & Restart to save and restart the app.6. Test#
If you receive a normal response after entering a message, the configuration was successful.Windows#
Windows 10/11 requires Git to be installed first.
Go to the Git official website, download the installer, then follow the prompts to finish the installation.1. Download the Claude App#
The download page requires access to the international internet.
2. Install After Downloading#
After installation is complete, the app page will open automatically.3. Enable Developer Mode#
Click the three-line menu in the top-left corner, choose Help -> Troubleshooting -> Enable Developer Mode, then select Enable in the pop-up.4. Open Third-Party Inference Configuration#
The app will restart automatically. Then click the three-line menu in the top-left corner and choose Developer -> Configure Third-Party Inference to open the configuration page.5. Fill in the Gateway Base URL and Gateway API Key#
On the Connection page, enter the Gateway base URL and Gateway API key, then set Gateway auth scheme to x-api-key.6. Add Models#
Scroll down to the Model list section, click Add to add a model, and enter a Model ID, for example:Click Apply locally at the bottom, then save. The app will restart automatically and enter the usage page.7. Test#
If you receive a normal response after entering a message, the configuration was successful.