India’s first GenAI unicorn is changing direction—and it’s raising big questions about the future of artificial intelligence startups. Krutrim, once positioned as a bold challenger to global AI leaders, is pivoting from building large AI models to focusing on cloud services. This shift reflects a growing reality in the AI industry: developing advanced models is expensive, complex, and increasingly dominated by global giants. So what does this move mean for India’s AI ambitions and the broader startup ecosystem?
 |
| Credit: Aparna Jayakumar/Bloomberg / Getty Images |
Krutrim AI Shift Marks a Turning Point in India’s GenAI Race
The decision by Krutrim to prioritize cloud infrastructure over in-house AI model development is more than just a business adjustment—it’s a signal of how challenging the generative AI landscape has become. After months of limited product updates and reduced public visibility, the company confirmed it is reallocating resources toward cloud-based AI services.
This strategic shift follows a broader restructuring effort that began in late 2025. The company paused its chip design initiatives, reassigned talent, and tightened its operational focus. These moves highlight the intense capital requirements needed to compete in AI model development, where only a few global players can sustain the scale required.
For a startup once seen as a national champion in AI, the pivot underscores a critical lesson: building foundational AI models is no longer just about innovation—it’s about financial endurance.
Why Building AI Models Is Becoming Increasingly Difficult
The challenges facing Krutrim are not unique. Across the global tech ecosystem, companies attempting to develop large-scale AI models are encountering similar barriers. Training advanced models requires enormous computing power, access to high-quality datasets, and sustained investment.
Global leaders like OpenAI, Google, and Anthropic have set a high benchmark, backed by billions in funding and vast infrastructure. Competing against such players is not just difficult—it’s often economically unsustainable for smaller startups.
This reality is pushing many companies to rethink their strategies. Instead of building models from scratch, they are focusing on applications, services, or infrastructure that leverage existing AI ecosystems. Krutrim’s move into cloud services fits squarely within this trend.
Krutrim’s Cloud Services Strategy Gains Momentum
While stepping back from model development might seem like a retreat, Krutrim is positioning its cloud business as a growth engine. The company reports increasing demand for its AI cloud services, with more than 25 enterprise customers across industries such as telecom, finance, and healthcare.
This shift reflects a broader market opportunity. As more businesses adopt AI, the need for reliable infrastructure—compute power, storage, and deployment tools—is growing rapidly. By focusing on this segment, Krutrim is tapping into a more immediate and scalable revenue stream.
Interestingly, the company also noted that most of its GPU compute capacity is already committed to external workloads. This suggests strong demand and hints at the potential profitability of infrastructure-focused AI businesses, especially in emerging markets like India.
Revenue Growth and Profitability Raise New Questions
Despite the strategic pivot, Krutrim has reported impressive financial growth. The company generated approximately ₹3 billion (around $31.5 million) in revenue in the financial year 2026, marking a threefold increase from the previous year. It also achieved its first annual net profit, with margins exceeding 10%.
However, these numbers come with important caveats. The company has not disclosed how much of its revenue comes from external customers versus its parent ecosystem. Earlier reports suggested that a significant portion of its earnings was tied to affiliated businesses.
This lack of transparency has led analysts to urge caution. Industry experts argue that profitability claims should be closely examined, especially in a sector where accounting complexities and internal dependencies can obscure the true financial picture.
Layoffs and Product Pullbacks Reflect Strategic Reset
Krutrim’s transformation hasn’t been without consequences. Over the past year, the company has reportedly cut more than 200 jobs across multiple rounds of layoffs. It also removed its AI assistant app from app stores, signaling a clear departure from its earlier product strategy.
These decisions reflect a broader reset. Instead of spreading resources across multiple initiatives, the company appears to be concentrating its efforts on areas with clearer commercial potential.
While layoffs are often viewed negatively, they are also a common feature of tech industry pivots. In this case, they highlight the difficult trade-offs startups must make when adapting to rapidly changing market conditions.
Rising Competition from Domestic and Global Players
As Krutrim shifts its focus, competitors are moving aggressively in the opposite direction. Indian rival Sarvam AI has continued to invest heavily in AI model development, showcasing new open-source models, hardware innovations, and strategic partnerships.
At the same time, global players are expanding their presence in India’s AI ecosystem. Companies like OpenAI and Google are actively engaging with local developers, enterprises, and policymakers, further intensifying competition.
This dynamic creates a challenging environment for domestic startups. They must navigate not only technological hurdles but also competitive pressures from well-funded international firms.
The Role of Leadership in Shaping Krutrim’s Future
The company’s direction is closely tied to its founder, Bhavish Aggarwal, who also leads Ola and Ola Electric. Known for his ambitious vision, Aggarwal initially positioned Krutrim as a cornerstone of India’s AI independence.
The pivot to cloud services suggests a more pragmatic approach. Rather than pursuing long-term ambitions without immediate returns, the company is focusing on sustainable growth and market demand.
Leadership decisions will play a crucial role in determining whether this strategy succeeds. Balancing ambition with practicality is often the difference between startup success and failure.
What This Means for India’s AI Ecosystem
Krutrim’s shift has broader implications for India’s AI landscape. It highlights the need for realistic expectations about what domestic startups can achieve in a globalized, capital-intensive industry.
At the same time, it opens up new opportunities. By focusing on infrastructure, companies can support the growing demand for AI adoption without directly competing in the most resource-intensive segments.
This could lead to a more balanced ecosystem, where different players specialize in various layers of the AI stack—from infrastructure to applications to research.
A Global AI Reality Check
The story of Krutrim is part of a larger global trend. As the initial hype around generative AI gives way to practical considerations, companies are reassessing their strategies.
Building cutting-edge AI models remains a long-term goal for many, but the path to profitability often lies elsewhere. Infrastructure, enterprise solutions, and applied AI are emerging as more viable business models in the near term.
For investors and entrepreneurs, this shift serves as a reminder that innovation must be paired with economic sustainability. The most successful companies will be those that can adapt quickly while maintaining a clear focus on value creation.
Krutrim’s pivot from AI model development to cloud services marks a defining moment for both the company and India’s AI ambitions. It reflects the harsh realities of competing in a global AI race dominated by a few powerful players, while also highlighting new opportunities in infrastructure and enterprise services.
As the AI industry continues to evolve, adaptability will be key. Krutrim’s journey shows that success in this space is not just about building the most advanced technology—it’s about finding the right strategy at the right time.
