Sarvam AI, India's fastest-growing artificial intelligence startup, has acknowledged that while India produces world-class AI talent, the country lacks experience in building frontier AI models from scratch. The admission came as the startup, recently valued at over $1 billion following HCLTech's Rs 1,427 crore investment, outlined its strategy for competing in the global AI race.
The Frontier Model Gap
Speaking at a industry event, Sarvam AI's leadership noted that India has an abundance of AI researchers and engineers skilled in applying AI to real-world problems, but very few who have been involved in training large-scale frontier models from scratch. This gap reflects a broader challenge for India's AI ecosystem — the country has world-class AI talent but limited experience at the very frontier of AI research where models like GPT-4, Claude, and Gemini are built.
Sarvam AI's approach is to bridge this gap by investing heavily in research and development, collaborating with global AI labs, and building domain-specific models that can compete in Indian and global markets without requiring the same scale of compute resources as general-purpose frontier models.
India's AI Ecosystem: Strengths and Weaknesses
| Dimension | India's Position | Global Benchmark |
|---|---|---|
| AI research papers | 3rd globally (by volume) | Behind US, China |
| AI talent pool | 2nd largest globally | Behind US |
| Frontier model experience | Very limited | US (OpenAI, Google, Anthropic) |
| AI compute infrastructure | Growing (India AI Mission) | Far behind US, China |
| AI startup funding | $5B+ in 2025-26 | Growing but still early |
| Enterprise AI adoption | Strong (IT services) | Comparable globally |
What This Means for Indian AI
Sarvam AI's candid assessment reflects a maturing understanding within India's tech ecosystem. While Indian IT services companies have been early adopters of AI tools, building foundational AI models requires rare expertise, massive compute infrastructure, and sustained research investment — all of which are still developing in India.
The Indian government's India AI Mission, with its focus on building compute infrastructure and funding AI research, aims to address some of these gaps. However, industry leaders acknowledge that catching up with frontier AI development will take years of sustained investment and talent development.
For Indian startups, the pragmatic path may be to focus on applied AI — building specialized models for India's unique market needs — while gradually developing the capability to compete at the frontier. Sarvam AI's strategy reflects this approach.
