Reliance Industries is preparing one of India’s largest artificial-intelligence infrastructure projects in Jamnagar, where a new generation of data centres, advanced Nvidia processors and renewable energy systems are being assembled into what the company describes as a sovereign AI backbone for the country.
Speaking at Reliance Industries’ 49th annual general meeting, Akash Ambani, the managing director of Jio Platforms, said Reliance Intelligence would commission the first 120 megawatts of AI compute capacity by the end of 2026.
The initial system will include a fleet of advanced Nvidia GB300 GPUs, with computing power that Reliance says will be equivalent to more than 75,000 H100 GPUs on an AI-inference basis. Once the first 120-megawatt phase becomes fully operational, Ambani said, capacity could rise to the equivalent of more than 200,000 H100 GPUs.
“This capacity places Reliance among the largest AI infrastructure platforms being built anywhere in the world,” Ambani said. “When compute becomes affordable, innovation becomes inevitable.”
The project is not simply a data-centre expansion. It reflects a larger effort by Reliance to place itself at the centre of India’s emerging AI economy — controlling infrastructure, delivering consumer applications, working with global technology companies and building systems designed to operate under Indian governance and data priorities.
Jamnagar Becomes the Centre of Reliance’s AI Ambition
Reliance has spent decades turning Jamnagar into an industrial centre associated with refining, petrochemicals and energy. It is now attempting to give the city another strategic role: that of a national-scale AI computing hub.
The company said the Jamnagar AI infrastructure would be powered entirely by clean energy generated through Reliance’s renewable platform in Kutch. That connection between high-performance computing and renewable power is significant because AI data centres require enormous quantities of electricity and cooling.
Globally, the rapid expansion of generative AI has intensified concerns over the environmental and energy costs of computing. Training and running large models can require thousands of processors operating continuously, pushing technology companies to secure long-term power supplies and build data centres near dependable sources of electricity.
Reliance’s advantage lies in its ability to connect several businesses that most technology companies must source separately: energy generation, telecommunications networks, cloud infrastructure, consumer distribution and enterprise services.
The company launched Reliance Intelligence in 2025 as a wholly owned deep-technology and artificial-intelligence subsidiary. Its first priority, Ambani said, would be to address what he called the greatest obstacle to AI adoption in India: the scarcity and high cost of computing capacity.
That challenge is particularly acute for Indian startups, universities, government institutions and smaller companies that cannot afford extensive access to advanced AI chips or foreign cloud platforms. By building domestic capacity at scale, Reliance is betting that it can reduce the cost of AI inference and make sophisticated models more widely available.
But the scale of the project also raises questions about market concentration. If a limited number of companies control the infrastructure required to train and operate advanced AI systems, they may gain substantial influence over pricing, access and the direction of innovation.
What Sovereign AI Means for Reliance
The term “sovereign AI” has become increasingly prominent as governments seek greater control over the data, infrastructure and models that underpin their digital economies.
In Reliance’s interpretation, sovereignty does not appear to mean technological isolation. Instead, the company is attempting to combine global technology with Indian infrastructure, execution, domain knowledge and governance.
“The right path is to combine the best global technologies with Indian execution, Indian infrastructure, Indian domain knowledge and India-first governance,” Ambani said.
That strategy is visible in Reliance’s partnerships with leading American technology companies.
Jio users already have access to Google AI Pro, powered by Gemini, while Reliance has also formed a joint venture with Meta intended to operationalise the company’s Llama open-source AI models for Indian enterprises.
The reliance on Nvidia hardware, Google models and Meta technology illustrates the complexity of sovereignty in the AI era. India may host the infrastructure, control deployment and shape local applications, while key components of the underlying technology remain developed abroad.
Reliance appears to be addressing this tension by focusing on control over deployment and delivery. Data can remain within Indian systems, applications can be tailored to Indian languages and industries, and governance can be designed around domestic regulatory and commercial requirements.
That approach may prove attractive to banks, government departments, healthcare providers and other institutions that handle sensitive information and are reluctant to rely entirely on foreign cloud infrastructure.
Still, sovereign AI will require more than domestic data centres. It will also depend on local research, indigenous intellectual property, skilled engineers, robust cybersecurity and the ability to maintain systems without excessive dependence on external vendors.
From Compute Infrastructure to Everyday AI Services
Reliance is not building AI infrastructure solely for researchers or large companies. Its ambition extends to consumer and small-business applications delivered through Jio’s enormous distribution network.
The company has introduced a collection of India-focused services designed to operate across 22 Indian languages.
These include JioBharatIQ, which Reliance says will make AI accessible to ordinary users; AI Vyapar, aimed at helping small merchants and businesses improve productivity; JioHealthIQ, designed to provide intelligent healthcare support; JioLearnIQ, focused on personalised education; and JioKrishiIQ, intended to help farmers make decisions about crops, weather, resources and income.
The applications reflect a broader shift in the AI industry. The early competition centred on building the most capable general-purpose model. The next phase is increasingly about distribution: placing AI inside familiar consumer services, business systems and sector-specific workflows.
Reliance possesses an advantage few AI companies can match. Jio already serves hundreds of millions of users through telecom, broadband, digital payments, entertainment and enterprise services. That network could allow Reliance to distribute AI tools at a scale that specialised technology startups would struggle to achieve.
It also gives the company access to an enormous range of potential use cases, from customer service and retail to agriculture and healthcare.
But widespread AI deployment creates significant governance challenges. Multilingual systems must be accurate across languages and dialects. Healthcare tools must avoid unsafe recommendations. Agricultural systems must account for local conditions. Business tools must protect confidential data. Consumer-facing products must clearly distinguish automated output from verified information.
The credibility of Reliance’s AI ambitions may therefore depend as much on safeguards and reliability as on computing power.
India’s AI Race Moves From Models to Infrastructure
Reliance’s Jamnagar project arrives as India seeks a larger role in the global AI economy.
Much of the early AI boom was dominated by American and Chinese companies that possessed access to advanced chips, large computing clusters and vast research budgets. Indian companies often built applications on top of foreign models rather than developing or hosting frontier systems themselves.
The emerging infrastructure push is an effort to change that balance.
A large domestic compute platform could support Indian startups, research institutions, enterprises and government agencies. It could make it easier to build models trained on Indian languages and datasets, develop industry-specific applications and operate sensitive systems without transferring data abroad.
Reliance’s scale makes it a powerful participant in that transition. But it will not be alone. Indian cloud providers, data-centre companies, startups and government-backed initiatives are also expanding AI capacity.
The central question is whether this investment will produce a broader ecosystem or simply create another layer of dependency, this time on a domestic conglomerate.
Affordable access, transparent pricing and support for independent developers will be critical. So will interoperability, competition and rules governing data use.
Reliance has framed the project as a national platform rather than merely a corporate asset. Its ability to fulfil that promise will depend on who gains access, on what terms and for what purposes.
For now, the company is making a straightforward bet: that the future of AI in India will be decided not only by who builds the best model, but by who controls the computing power needed to run it.
In Jamnagar, Reliance is assembling that power on a scale intended to serve hundreds of millions of users — and to place India more firmly inside the global contest over artificial intelligence.
