OSDHack 2026

OSDHack 2026 Build AI that runs closer to the user - faster, lighter, more private, and open source. OSDHack 2026 is an online open-source hackathon organized by Open Source Developer’s Community, an independent open-source community. This year’s theme is On Device AI. Build a project where the core AI feature runs locally on a phone, laptop, desktop, browser, edge device, or embedded system. Think local models, offline-first AI, privacy-friendly tools, browser-based AI, lightweight inference, edge AI, embedded AI, and creative open-source projects that do not fully depend on cloud AI APIs for their main AI functionality. Theme: On Device AI Your project should focus on On Device AI. You can build around: Local AI models Browser-based AI Mobile on-device AI Desktop or laptop AI tools Edge or embedded AI Offline-first AI apps Privacy-focused AI utilities Lightweight local inference Cloud services are allowed for support features like hosting, authentication, storage, databases, or deployment. However, the main AI feature must run on-device, locally, in-browser, on edge, or on embedded hardware. In short: cloud support is okay, but the core AI magic should happen locally. Timeline Event Date and Time Hackathon Starts 10 July 2026, 6:00 PM IST Hackathon Ends 15 July 2026, 6:00 PM IST Final Submission Deadline 15 July 2026, 6:00 PM IST Evaluation Period TBD Results Announcement TBD Judging Criteria TBD Prize Details TBD All deadlines follow Indian Standard Time (IST). Hackathon Round There are no fixed problem statements. Pick your own idea and build something related to On Device AI. The project must be built during the hackathon period: 10 July 2026, 6:00 PM IST to 15 July 2026, 6:00 PM IST Only work committed before 15 July 2026, 6:00 PM IST will be considered for judging. Later commits may remain in the repository, but they will not be evaluated. What to Submit Submit your project on Unstop before the deadline. Your submission should make it easy to understand: What you built Why it matters How it works How it uses On Device AI How others can run or try it Your need to submit a public git repository link, it should include: Source code Clear README OSI-compliant open-source license Setup or usage instructions Demo video link Screenshots You may use GitHub, GitLab, Codeberg, Bitbucket, or any other public Git hosting platform. Open Source Requirement OSDHack 2026 is an open-source hackathon. Your project must be fully open source and use an OSI-compliant license. You may use open-source libraries, models, datasets, frameworks, templates, and tools. Just follow their licenses and give proper attribution wherever required. Do not use copied private code, stolen assets, or proprietary material without permission. New to On Device AI? No worries - you do not need to be an AI expert to participate. On Device AI simply means the AI runs directly on the user’s device instead of relying entirely on a remote cloud AI service. That device could be: A browser A laptop or desktop A phone An edge device An embedded board or microcontroller You also do not need expensive hardware or a powerful GPU to get started. A simple browser-based, desktop, or mobile project can be a great submission. Where Can You Start? There is no single “correct” way to build an On Device AI project. Start with something small, fun, and useful. You could build: A browser tool that works offline A local AI assistant A privacy-friendly productivity app A smart camera or detection system A lightweight chatbot A speech or audio utility A document helper A TinyML or robotics experiment An AI-powered accessibility tool A creative AI experience that runs locally The goal is not to build the biggest model possible, it is to build something thoughtful, practical, creative, or fun using local AI. Helpful Resources Here are some beginner-friendly resources to explore: Browser and Web AI Here: Transformers.js: https://huggingface.co/docs/transformers.js ONNX Runtime Web: https://onnxruntime.ai/docs/tutorials/web/ WebLLM: https://webllm.mlc.ai/docs/ Great for browser-based AI apps, local inference, and privacy-friendly web experiences. Mobile AI Here: LiteRT / TensorFlow Lite: https://ai.google.dev/edge/litert ONNX Runtime Mobile: https://onnxruntime.ai/docs/get-started/with-mobile.html MediaPipe: https://ai.google.dev/edge/mediapipe/solutions/guide ExecuTorch: https://docs.pytorch.org/executorch/stable/index.html Useful for camera apps, speech tools, offline mobile utilities, and real-time AI experiences.