Compute & Cloud
Krutrim pivots to cloud services as AI model ambitions cool
India’s Krutrim is shifting from AI model development to cloud services following a business overhaul, layoffs, and a pivot toward more viable infrastructure-focused revenue streams.
On Tuesday, Krutrim said it was moving toward cloud services, shifting its focus from AI model development. The transition follows a business overhaul in late 2025, which included reallocating capital and talent, as well as pausing its chip design efforts. The strategic shift comes after the Bengaluru-based startup released its base model (core AI model architecture). Krutrim had initially positioned itself as one of India’s earliest GenAI (Generative Artificial Intelligence) contenders, aiming to build domestic alternatives to models from competitors like OpenAI, Anthropic, Google, and xAI. However, Krutrim has experienced a period of limited public activity, including pulling its Kruti AI assistant app from app stores in April, with its last post on X dating back to December. Meanwhile, rival Sarvam participated in the AI Impact Summit in New Delhi, showcasing open-source models and hardware developments, and partnering with space-tech firm Pixxel.
The restructuring has also impacted Krutrim’s workforce. According to local media reports, the startup cut more than 200 roles across multiple rounds of layoffs. Despite these challenges, Krutrim reported its financial performance for the financial year 2026, though questions remain regarding its revenue mix. Previously, in the financial year 2025, about 90% of its revenue came from group companies within the ecosystem of parent company Ola, which is led alongside EV maker Ola Electric by Krutrim founder Bhavish Aggarwal.
The company’s reported financial metrics for the financial year 2026 include:
- Revenue: About ₹3 billion (around $31.52 million).
- Profitability: Its first annual net profit, with margins exceeding 10%.
- Customer Base: More than 25 enterprise customers across sectors including telecom, financial services, and healthcare.
Krutrim stated that most of its GPU compute capacity (computing power provided by Graphics Processing Units) is already committed to external workloads, pointing to growing demand for its AI cloud services. However, analysts caution that the company’s profitability claims require verification. Sanchit Vir Gogia, chief analyst at Greyhound Research, noted that while the move toward cloud services is commercially sensible, the startup’s financial claims must be tested. “The standard of proof must rise with the claim,” Gogia said. This caution comes after Krutrim raised $50 million at a $1 billion valuation in January 2024.
Why it matters
Krutrim’s pivot highlights the challenging economics of building large-scale AI systems, forcing a shift from ambitious model development to more immediate, infrastructure-focused revenue streams.