Meta has agreed to lease a 168-megawatt, artificial intelligence-ready data centre in India from Mukesh Ambani’s Reliance Industries, its first such facility in the country. Reliance will build the site in Jamnagar, Gujarat, and Meta will lease it with an option to scale, the companies said on Tuesday, June 10, 2026. The announcement marks a major step in Meta’s global infrastructure strategy as it expands compute capacity outside the United States and Europe. The Jamnagar campus is positioned near undersea cable landings and Reliance Jio’s fibre network, giving Meta low-latency access to a billion-plus user base. The deal deepens a partnership that began with Meta’s 5.7 billion dollar investment in Reliance’s Jio Platforms in 2020, followed by a joint venture to bring Meta’s open-source artificial intelligence models to Indian businesses.
“This world-class facility in Jamnagar will help us scale our artificial intelligence infrastructure globally while deepening our long-term investment in India’s economy,” Meta chief executive Mark Zuckerberg said. Meta frames the site as part of its push toward what it calls “personal superintelligence”. The data centre will run on renewable energy and use desalinated seawater for cooling, with Meta covering the full cost of the power and water. The 168 megawatt capacity is a first phase, with room to grow as Meta’s model training and inference workloads increase. Jamnagar is where Reliance is building one of the world’s largest data centre campuses, leveraging land availability, grid access, and proximity to industrial infrastructure.
The deal lands amid a wider rush of hyperscalers to India. More than 400 billion dollars flowed into India’s artificial intelligence ecosystem over the past year, most of it into data centres and the power to run them. Meta runs vast compute in the United States, from a 200 billion dollar Louisiana campus to a 21 billion dollar CoreWeave deal, but India offers cheaper land, available power, and a massive user base. The pull is partly a push: United States and European grids are straining under artificial intelligence load, with utilities planning 1.4 trillion dollars of spending, while India’s capacity is booming. Hyperscale operators are seeking markets that can deliver power, water, and permitting at speed, and Gujarat has positioned itself as a preferred destination for data centre infrastructure in India.
To power the buildout cleanly, Meta also contracted nearly 1 gigawatt of Indian renewables, 837 megawatts from CleanMax and 88 megawatts from Fourth Partner Energy. The open question is whether that keeps pace. Running a 168 megawatt site, with room to grow, on renewables and desalinated water is the pitch, but India’s grid and water systems are under their own strain. Whether Jamnagar stays green as Meta scales it is the test ahead. Desalination addresses water scarcity but is energy intensive, making the renewable power contracts critical to the sustainability claim. The integration of power, water, and compute in a single campus reflects a new model for artificial intelligence infrastructure where energy sourcing is as strategic as server architecture.
Jamnagar’s advantages include industrial-scale power availability, land for expansion, and coastal access for seawater cooling. Reliance will build and run the site, allowing Meta to focus on hardware deployment and workload optimization. The arrangement reduces Meta’s construction risk while giving Reliance a long-term anchor tenant for its data centre platform. For Reliance, the deal validates its pivot from energy to digital infrastructure and leverages its existing investments in solar, wind, and green hydrogen. For Meta, it secures sovereign-scale compute in a growth market without the multi-year timelines typical of United States grid upgrades. The site will support Meta’s open-source model ecosystem, including Llama family models used by Indian enterprises through the Meta Reliance joint venture.
India’s policy environment has encouraged data centre investment through state incentives, single-window clearances, and renewable energy open access rules. Gujarat’s data centre policy and Jamnagar’s industrial zoning reduce development friction compared to saturated markets. The proximity to undersea cables improves latency for South Asia, Middle East, and East Africa traffic. As artificial intelligence workloads shift from training to inference, geographic distribution becomes vital for user experience and data compliance. Meta’s lease model also provides flexibility to scale capacity in line with product roadmaps, rather than overbuilding upfront. Industry analysts view the move as a template for how hyperscalers will expand in power-rich emerging markets while managing sustainability commitments. Details on global data centre sustainability trends can be found via the Uptime Institute.
With the Jamnagar facility, Meta joins other hyperscalers sending artificial intelligence compute to India to balance cost, power, and market access. The 168 megawatt first phase establishes a footprint that can expand to multi-hundred megawatt scale if demand holds. The success of the project will depend on timely renewable delivery, reliable desalination operations, and stable grid integration. If executed, it strengthens India’s position as a global artificial intelligence infrastructure hub and deepens the Reliance Meta alliance across connectivity, models, and compute. The deal signals that the next wave of artificial intelligence growth will be built where power and policy align, and Gujarat is positioning itself at the centre of that map.
