Project Mosaic launched Version One of India’s first decentralized AI grid, attempting to replace expensive centralized data centers by distributing compute using AIKosh IndiaAI datasets.
Noida (Uttar Pradesh) [India], August 25 — Project Mosaic has officially launched Version One of its groundbreaking web-based AI network. Rather than relying on large, capital-intensive cloud data centers, the platform aggregates everyday consumer laptops and mobile phones to collaboratively train massive 100-billion-parameter AI models. This architecture shifts high-performance computing power directly to end users.
“Right now, the entire AI industry is hitting a massive wall: building giant multi-billion-dollar data centers that consume city-scale electricity, water, and expose everyone’s private data to the cloud,” said Kanishk Anand, Founder of Project Mosaic. “Project Mosaic replaces the need for those centralized data centers by aggregating the millions of powerful GPUs and NPUs already sitting idle on everyday laptops, workstations, and phones. Using browser-native WebGPU and Federated Learning, we distribute AI training across the edge. Zero server energy bills, 100% zero-leakage privacy, and infinite scalability powered by the crowd.”
The launch aligns with Prime Minister Narendra Modi’s Independence Day address on August 15, which outlined a national objective to train one crore youth in artificial intelligence skills within the coming year. Project Mosaic supports this initiative by providing free, highly accessible AI training and developing its own Small Language Model. Historically, prohibitive cloud server costs excluded students from AI development. By converting standard laptops into active AI nodes, the platform removes these financial barriers, offering a grassroots pathway for millions of urban and rural Indian students to transition from passive consumers to active technology builders.
By harnessing the latent computing capacity of India’s 1.4 billion citizens, Project Mosaic is cultivating a self-reliant (Aatmanirbhar) digital ecosystem. While many global AI models are trained on Western datasets, they often overlook India’s linguistic and cultural nuances. To address this, Project Mosaic natively integrates with the government’s AIKosh IndiaAI Datasets Platform and Indian Government Repositories. This ensures models trained on the decentralized network accurately comprehend Indian languages and heritage. Crucially, because all processing occurs locally on user devices, sensitive domestic data never leaves the country or enters foreign cloud servers.
Traditional AI data centers present significant ecological challenges, consuming megawatts of electricity and millions of gallons of water. Project Mosaic mitigates this infrastructure bottleneck through a more sustainable model. However, purely local AI can struggle with complex queries requiring real-time internet data. To bridge this gap, the platform features EcoMind, a proprietary, ultra-efficient Green AI search engine embedded directly into the ecosystem. When a local model encounters a demanding query, EcoMind acts as a sustainable backup, securely fetching verified information from the web without compromising device performance.
Building the future of AI does not require software engineering expertise. Project Mosaic has gamified participation through its Arcade Hub, where users can play retro-style web games such as Neural Super-Highway Racer, Sky Shooter, and Quantum Snakes & Ladders. While playing, the browser quietly utilizes the device’s graphics chip to solve minor AI mathematics tasks in the background. No downloads, driver updates, or coding knowledge are required—users simply engage with the games to contribute to the national computing grid.
Instead of rewriting the massive AI brain on a single computer, each device processes a microscopic fragment of data, creating a tiny mathematical “sticky note” of what it learned. Before this note leaves the computer, a mathematical scrambling filter ensures no one can ever read the raw data. A central system securely collects these notes from thousands of users simultaneously, discards any errors, and blends them into one master AI upgrade that is instantly beamed back to everyone on the network.
As India’s pioneer in decentralized, browser-native AI infrastructure, Project Mosaic is dedicated to eliminating the cost, environmental, and privacy constraints of centralized AI systems. The platform enables consumer hardware to collaboratively develop foundation models securely at the edge.
If you object to the content of this press release, please notify us at [email protected]. We will respond and rectify the situation within 24 hours.