How to Set Up and Use Gemini with Janitor AI (Complete 2026 Tutorial)

Open your preferred web browser and launch a new tab. Type "Janitor AI" in the search bar and press Enter to locate the official platform. Click the official Janitor AI website from the search results.

Once the Janitor AI homepage loads, sign in to your account. If you aren't already logged in, enter your credentials to access your dashboard. This ensures you have full access to the configuration settings you'll modify throughout this tutorial.

Your Janitor AI interface should now display your existing chats or character options. You're ready to proceed with API configuration. The video covers this section from the 13-second mark through approximately 30 seconds [00:13 - 00:30].

From your Janitor AI dashboard, open any existing chat or create a new conversation. This brings you into the main chat interface where configuration options are located. You don't need to send any messages—opening the chat provides access to the settings menu.

Look to the top right corner of the screen for the three horizontal lines icon (the hamburger menu). Click this menu icon to reveal a dropdown list of available options. This menu contains various settings and configuration tools for your account.

Locate and click on API Settings from the dropdown menu. This option takes you to the configuration page where you manage all API connections. The video demonstrates this navigation between [00:31 - 00:40].

Inside the API Settings page, you'll see multiple tabs for different connection types. Click on the Proxy tab to access the proxy configuration interface. This is where you set up the Gemini connection using a proxy server rather than a direct API integration.

The proxy method allows flexible routing of your API requests and provides additional features. The proxy tab contains fields for managing multiple configurations, allowing you to maintain different setups for various AI models or use cases.

The interface displays any existing proxy configurations you've created. If this is your first setup, the area will be empty. This section corresponds to the [00:40 - 00:55] portion of the video.

Within the Proxy tab, locate and click the Add Configuration button. This initiates creation of a new proxy setup for your Gemini integration. A new configuration form will appear with several empty fields to populate.

The first field is the configuration name. Enter a clear, descriptive name such as "Gemini 2.5" or "Google Gemini Proxy" so you can easily identify this configuration later. Memorable naming is particularly important if you maintain multiple API configurations for different models.

This naming convention helps you quickly switch between different AI backends when chatting. Good naming also makes troubleshooting easier if you need to revisit settings. The video covers this step between [00:40 - 00:55].

In the Model Name field, type the exact model identifier for Gemini: gemini-2.5-flash-preview-05-20 precisely as shown, including all spaces and hyphens. This exact string tells the proxy server which specific Gemini model variant to route your requests to.

Pay close attention to capitalization, spacing, and punctuation. Even a single typo or extra space causes integration failure with a "model not found" error. The system requires exact matching for model identifiers to ensure proper routing.

Double-check your entry before moving to the next field. Compare character-by-character against the provided model name to ensure accuracy. This critical step is demonstrated in the video from approximately [00:55 - 01:31].

Locate the Proxy URL field in your configuration form. This field requires the full HTTPS address of the proxy server that handles communication between Janitor AI and Google's Gemini API. Paste the complete proxy URL provided to you for this integration.

The proxy URL should begin with "https://" and contain the full domain and any necessary path components. Copy the entire URL without any trailing spaces, line breaks, or additional punctuation, as these characters commonly cause connection errors.

After pasting, visually inspect both the beginning and end of the URL to confirm nothing was accidentally included or cut off. The proxy URL acts as the bridge between Janitor AI and Gemini, so accuracy is essential. This step continues the section from [00:55 - 01:31].

Open a new browser tab while keeping your Janitor AI configuration page open. Navigate to Google AI Studio by typing it into your search bar or entering the direct URL. This is Google's official interface for managing API access to their AI models including Gemini.

Once Google AI Studio loads, look for the Get API Key option on the lower left side of the interface. Click this button to navigate to the API key management page. This area allows you to create, view, and manage all your Google AI API credentials.

On the API keys page, locate the Create API Key button in the top right corner of the screen. Click this button to begin the key generation process. You may be prompted to select a specific Google Cloud project if you have multiple projects associated with your account.

If a project selection dialog appears, choose the appropriate project for your Gemini integration or select the default option. Click Create Key to confirm your selection and generate the new API key. After a brief moment, your new API key will appear on the screen.

Click the copy icon next to your newly generated API key to copy it to your clipboard. This key is a sensitive credential that grants access to your Google AI services—treat it like a password and never share it publicly. The video walks through this entire process from [01:31 - 02:08].

Return to your Janitor AI browser tab where you left the proxy configuration form open. Locate the API Key field in the configuration form. Click inside this field and paste the Google API key you just copied from Google AI Studio.

Verify that the entire API key has been pasted correctly without extra spaces at the beginning or end. The key is a long string of characters, and every character must be included for authentication to succeed. Visually confirm the paste operation was complete.

Click the Add Configuration button at the bottom of the form to register this new proxy setup. The system saves your configuration with the name, model identifier, proxy URL, and API key you provided. Your new configuration should appear in the list of available proxy setups.

Finally, click the Save Settings button to apply all changes to your Janitor AI account. This commits your configuration and makes it available for use in your chats. The video demonstrates this finalization from [02:08 - 02:24].

Authentication errors after setup typically indicate an incorrect or expired API key. Return to API Settings and verify that your Google API key is pasted completely and accurately. Generate a new key in Google AI Studio if needed and replace the old one.

Model not found errors usually indicate a typo in the model name field. Navigate back to your proxy configuration and carefully check that gemini-2.5-flash-preview-05-20 is spelled exactly as shown. Compare each character and space to ensure perfect accuracy.

Connection timeout issues often stem from an incorrect or inaccessible proxy URL. Verify that the proxy address is entered correctly and that your network can reach the proxy server. Some corporate or school networks may block certain proxy connections, so you might need to try from a different network.

If messages fail to send or you receive generic error responses, try refreshing your browser and reopening the chat. Clearing your browser cache can resolve unexpected behavior. You can also temporarily switch to a different API configuration to determine if the issue is specific to your Gemini setup.

You can always return to API Settings to edit, update, or delete your configuration. The system allows you to maintain multiple API setups and switch between them as needed. This troubleshooting section corresponds to the video from [02:24 - 02:50].

With your configuration saved and tested, you're ready to use Google Gemini within Janitor AI. Open any character chat or start a new conversation to experience Gemini's capabilities. Your messages are now processed through the Gemini model via the proxy connection you configured.

You should notice faster response times compared to other models, as Gemini is optimized for speed. The model supports multimodal capabilities, meaning it handles various types of input and context. Your chat experience should feel smooth and responsive with the integration properly configured.

Return to API Settings at any time to modify your configuration, rotate your API key for security purposes, or add additional model setups. Managing multiple configurations allows you to switch between different AI models based on your needs. This concluding section covers the final portion of the video from approximately [02:50 - end].