The latest articles on DEV Community by Mattboxx (@mattboxx). https://dev.to/mattboxx https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4208322%2F919ea442-5569-401f-8dac-d34d80a3a63d.jpg https://dev.to/mattboxx en Mattboxx Mon, 26 Oct 2026 09:48:23 +0000 https://dev.to/mattboxx/battery-forecast-local-ai-battery-predictions-for-windows-3j83 https://dev.to/mattboxx/battery-forecast-local-ai-battery-predictions-for-windows-3j83 <p>I created Battery Forecast, an open-source Windows adaptation of Matteo Bonfanti’s Till Dusk, powered by Mistral AI’s Timestral-28M model.</p> <p>Instead of relying only on Windows’ battery estimates, it uses your local battery history to predict:</p> <ul> <li>How long your battery will last based on your usage.</li> <li>When charging will reach your target (e.g. 80%).</li> <li>Your expected battery level later in the day.</li> <li>Battery trends over the next 24 hours.</li> </ul> <p>Features:</p> <ul> <li>Fully offline AI inference — no cloud or account required.</li> <li>Battery history and health monitoring.</li> <li>System tray app with customizable alerts.</li> <li>Estimated time ranges and interactive charts.</li> <li>Supports Windows 10/11 (x64 and ARM64).</li> </ul> <p>The model runs locally on your CPU and uses historical data without retraining.</p> <p>Note: Forecast accuracy is still being evaluated, and charging predictions remain experimental.</p> <p>GitHub: <a href="https://github.com/Mattboxx/battery-forecast" rel="noopener noreferrer">https://github.com/Mattboxx/battery-forecast</a></p> <p>Website: <a href="https://mattboxx.github.io/battery-forecast/" rel="noopener noreferrer">https://mattboxx.github.io/battery-forecast/</a></p> <p>I’d love to hear your feedback, especially from people willing to test the predictions on their own laptops!</p> mistralai battery windows software