base on Control Any Computer Using LLMs. # Open Interface <picture> <img src="assets/icon.png" align="right" alt="Open Interface Logo" width="120" height="120"> </picture> ### Control Your Computer Using LLMs Open Interface - Self-drives your computer by sending your requests to an LLM backend (GPT-4o, Gemini, etc) to figure out the required steps. - Automatically executes these steps by simulating keyboard and mouse input. - Course-corrects by sending the LLM backend updated screenshots of the progress as needed. <div align="center"> <h4>Full Autopilot for All Computers Using LLMs</h4> [](https://github.com/AmberSahdev/Open-Interface?tab=readme-ov-file#install) [](https://github.com/AmberSahdev/Open-Interface?tab=readme-ov-file#install) [](https://github.com/AmberSahdev/Open-Interface?tab=readme-ov-file#install) <br> []((https://github.com/AmberSahdev/Open-Interface/releases/latest))    [](https://github.com/AmberSahdev/Open-Interface/releases/latest) </div> ### <ins>Demo</ins> 💻 "Solve Today's Wordle"<br> <br> *clipped, 2x* <details> <summary><a href="https://github.com/AmberSahdev/Open-Interface/blob/main/MEDIA.md#demos">More Demos</a></summary> <ul> <li> "Make me a meal plan in Google Docs" <img src="assets/meal_plan_demo_2x.gif" style="margin: 5px; border-radius: 10px;"> </li> <li> "Write a Web App" <img src="assets/code_web_app_demo_2x.gif" style="margin: 5px; border-radius: 10px;"> </li> </ul> </details> <hr> ### <ins>Install</ins> 💽 <details> <summary><img src="https://upload.wikimedia.org/wikipedia/commons/thumb/8/84/Apple_Computer_Logo_rainbow.svg/960px-Apple_Computer_Logo_rainbow.svg.png?20250629104313" alt="MacOS Logo" width="13" height="15"> <b>MacOS</b></summary> <ul> <li>Download the MacOS binary from the latest <a href="https://github.com/AmberSahdev/Open-Interface/releases/latest">release</a>.</li> <li>Unzip the file and move Open Interface to the Applications Folder.<br><br> <img src="assets/macos_unzip_move_to_applications.png" width="350" style="border-radius: 10px; border: 3px solid black;"> </li> </ul> <details> <summary><b>Apple Silicon M-Series Macs</b></summary> <ul> <li> Open Interface will ask you for Accessibility access to operate your keyboard and mouse for you, and Screen Recording access to take screenshots to assess its progress.<br> </li> <li> In case it doesn't, manually add these permission via <b>System Settings</b> -> <b>Privacy and Security</b> <br> <img src="assets/mac_m3_accessibility.png" width="400" style="margin: 5px; border-radius: 10px; border: 3px solid black;"><br> <img src="assets/mac_m3_screenrecording.png" width="400" style="margin: 5px; border-radius: 10px; border: 3px solid black;"> </li> </ul> </details> <details> <summary><b>Intel Macs</b></summary> <ul> <li> Launch the app from the Applications folder.<br> You might face the standard Mac <i>"Open Interface cannot be opened" error</i>.<br><br> <img src="assets/macos_unverified_developer.png" width="200" style="border-radius: 10px; border: 3px solid black;"><br> In that case, press <b><i><ins>"Cancel"</ins></i></b>.<br> Then go to <b>System Preferences -> Security and Privacy -> Open Anyway.</b><br><br> <img src="assets/macos_system_preferences.png" width="100" style="border-radius: 10px; border: 3px solid black;"> &nbsp; <img src="assets/macos_security.png" width="100" style="border-radius: 10px; border: 3px solid black;"> &nbsp; <img src="assets/macos_open_anyway.png" width="400" style="border-radius: 10px; border: 3px solid black;"> </li> <br> <li> Open Interface will also need Accessibility access to operate your keyboard and mouse for you, and Screen Recording access to take screenshots to assess its progress.<br><br> <img src="assets/macos_accessibility.png" width="400" style="margin: 5px; border-radius: 10px; border: 3px solid black;"><br> <img src="assets/macos_screen_recording.png" width="400" style="margin: 5px; border-radius: 10px; border: 3px solid black;"> </li> </ul> </details> <ul> <li>Lastly, checkout the <a href="#setup">Setup</a> section to connect Open Interface to LLMs (OpenAI GPT-4V)</li> </ul> </details> <details> <summary><img src="https://upload.wikimedia.org/wikipedia/commons/a/af/Tux.png" alt="Linux Logo" width="18" height="18"> <b>Linux</b></summary> <ul> <li>Linux binary has been tested on Ubuntu 20.04 so far.</li> <li>Download the Linux zip file from the latest <a href="https://github.com/AmberSahdev/Open-Interface/releases/latest">release</a>.</li> <li> Extract the executable and checkout the <a href="https://github.com/AmberSahdev/Open-Interface?tab=readme-ov-file#setup">Setup</a> section to connect Open Interface to LLMs, such as OpenAI GPT-4V.</li> </ul> </details> <details> <summary><img src="https://upload.wikimedia.org/wikipedia/commons/5/5f/Windows_logo_-_2012.svg" alt="Linux Logo" width="15" height="15"> <b>Windows</b></summary> <ul> <li>Windows binary has been tested on Windows 10.</li> <li>Download the Windows zip file from the latest <a href="https://github.com/AmberSahdev/Open-Interface/releases/latest">release</a>.</li> <li>Unzip the folder, move the exe to the desired location, double click to open, and voila.</li> <li>Checkout the <a href="https://github.com/AmberSahdev/Open-Interface?tab=readme-ov-file#setup">Setup</a> section to connect Open Interface to LLMs (OpenAI GPT-4V)</li> </ul> </details> <details> <summary><img src="https://upload.wikimedia.org/wikipedia/commons/thumb/c/c3/Python-logo-notext.svg/120px-Python-logo-notext.svg.png?20250701090410" alt="Python Logo" width="15" height="15"> <b>Run as a Script</b></summary> <ul> <li>Clone the repo <code>git clone https://github.com/AmberSahdev/Open-Interface.git</code></li> <li>Enter the directory <code>cd Open-Interface</code></li> <li><b>Optionally</b> use a Python virtual environment <ul> <li>Note: pyenv handles tkinter installation weirdly so you may have to debug for your own system yourself.</li> <li><code>pyenv local 3.12.2</code></li> <li><code>python -m venv .venv</code></li> <li><code>source .venv/bin/activate</code></li> </ul> </li> <li>Install dependencies <code>pip install -r requirements.txt</code></li> <li>Run the app using <code>python app/app.py</code></li> </ul> </details> ### <ins id="setup">Setup</ins> 🛠️ <details> <summary><b>Set up the OpenAI API key</b></summary> - Get your OpenAI API key - Open Interface needs access to GPT-4o to perform user requests. GPT-4o keys can be downloaded from your OpenAI account at [platform.openai.com/settings/organization/api-keys](https://platform.openai.com/settings/organization/api-keys). - [Follow the steps here](https://help.openai.com/en/articles/8264644-what-is-prepaid-billing) to add balance to your OpenAI account. To unlock GPT-4o a minimum payment of $5 is needed. - [More info](https://help.openai.com/en/articles/7102672-how-can-i-access-gpt-4) - Save the API key in Open Interface settings - In Open Interface, go to the Settings menu on the top right and enter the key you received from OpenAI into the text field like so: <br> <br> <picture> <img src="assets/set_openai_api_key.png" align="middle" alt="Set API key in settings" width="400"> </picture><br> <br> - After setting the API key for the first time you'll need to <b>restart the app</b>. </details> <details> <summary><b>Set up the Google Gemini API key</b></summary> - Go to Settings -> Advanced Settings and select the Gemini model you wish to use. - Get your Google Gemini API key from https://aistudio.google.com/app/apikey. - Save the API key in Open Interface settings. - Save the settings and <b>restart the app</b>. </details> <details> <summary><b>Optional: Setup a Custom LLM</b></summary> - Open Interface supports using other OpenAI API style LLMs (such as Llava) as a backend and can be configured easily in the Advanced Settings window. - Enter the custom base url and model name in the Advanced Settings window and the API key in the Settings window as needed. - NOTE - If you're using Llama: - You may need to enter a random string like "xxx" in the API key input box. - You may need to append /v1/ to the base URL. <br> <picture> <img src="assets/advanced_settings.png" align="middle" alt="Set API key in settings" width="400"> </picture><br> <br> - If your LLM does not support an OpenAI style API, you can use a library like [this](https://github.com/BerriAI/litellm) to convert it to one. - You will need to restart the app after these changes. </details> <hr> ### <ins>Stuff It’s Error-Prone At, For Now</ins> 😬 - Accurate spatial-reasoning and hence clicking buttons. - Keeping track of itself in tabular contexts, like Excel and Google Sheets, for similar reasons as stated above. - Navigating complex GUI-rich applications like Counter-Strike, Spotify, Garage Band, etc due to heavy reliance on cursor actions. ### <ins>The Future</ins> 🔮 (*with better models trained on video walkthroughs like Youtube tutorials*) - "Create a couple of bass samples for me in Garage Band for my latest project." - "Read this design document for a new feature, edit the code on Github, and submit it for review." - "Find my friends' music taste from Spotify and create a party playlist for tonight's event." - "Take the pictures from my Tahoe trip and make a White Lotus type montage in iMovie." ### <ins>Notes</ins> 📝 - Cost Estimation: $0.0005 - $0.002 per LLM request depending on the model used.<br> (User requests can require between two to a few dozen LLM backend calls depending on the request's complexity.) - You can interrupt the app anytime by pressing the Stop button, or by dragging your cursor to any of the screen corners. - Open Interface can only see your primary display when using multiple monitors. Therefore, if the cursor/focus is on a secondary screen, it might keep retrying the same actions as it is unable to see its progress. <hr> ### <ins>System Diagram</ins> 🖼️ ``` +----------------------------------------------------+ | App | | | | +-------+ | | | GUI | | | +-------+ | | ^ | | | | | v | | +-----------+ (Screenshot + Goal) +-----------+ | | | | --------------------> | | | | | Core | | LLM | | | | | <-------------------- | (GPT-4o) | | | +-----------+ (Instructions) +-----------+ | | | | | v | | +-------------+ | | | Interpreter | | | +-------------+ | | | | | v | | +-------------+ | | | Executer | | | +-------------+ | +----------------------------------------------------+ ``` --- ### <ins>Star History</ins> ⭐️ <picture> <img src="https://api.star-history.com/svg?repos=AmberSahdev/Open-Interface&type=Date" alt="Star History" width="720"> </picture> ### <ins>Links</ins> 🔗 - Check out more of my projects at [AmberSah.dev](https://AmberSah.dev). - Other demos and press kit can be found at [MEDIA.md](MEDIA.md). <div align="center"> <img alt="GitHub Repo stars" src="https://img.shields.io/github/stars/AmberSahdev/Open-Interface"> <a href="https://github.com/AmberSahdev"> <img alt="GitHub followers" src="https://img.shields.io/github/followers/AmberSahdev"> </a> </div> ", Assign "at most 3 tags" to the expected json: {"id":"12729","tags":[]} "only from the tags list I provide: []" returns me the "expected json"