Falling AI Token Costs a Win for India, says Zoho's Vembu
Zoho co-founder Sridhar Vembu says declining AI token costs will boost affordability and reduce corporate spending in India.
The rapid decline in artificial intelligence (AI) token costs is set to be a significant positive development for India, according to Zoho co-founder Sridhar Vembu. As the technology becomes more affordable, corporate spending on AI tools is expected to decrease in the coming years, making advanced intelligence more accessible across the country.
The Commoditization of Intelligence
Speaking at the launch of the World Development Report 2026, Vembu observed that intelligence is undergoing a process of rapid commoditization. Even as global corporations continue to pour massive investments into the technology, the cost of accessing these models is plummeting.
“Token costs are falling rapidly and intelligence is actually commoditizing rapidly…and that's a good trend for India,” Vembu stated.
The Software Development Feedback Loop
Vembu explained that the primary economic value of AI currently stems from software development. He detailed a specific four-step cycle that drives modern AI applications:
- Large Language Models (LLMs): The initial generation of software code.
- Code Generation: The transformation of prompts into functional programming.
- Verification Layer: Checking the generated code for accuracy.
- Testing/Execution Layer: Running the code to ensure it works as intended.
This continuous feedback loop allows AI systems to identify errors, correct themselves, and refine their output over time. Vembu noted that much of the current economic utility of AI is derived from this iterative process, often involving specialized software languages like Lean.
Massive Computing Power and Shifting Expenditures
The true strength of artificial intelligence lies in its immense computational capacity. While standard computers operate at scales of billions of computations, AI systems function at the scale of trillions, allowing them to learn from vast amounts of feedback.
The financial scale of AI adoption is staggering. Vembu highlighted the massive disparity in spending between tech giants and growing firms:
- Microsoft: Spends approximately USD 1 billion annually on AI for software development.
- Zoho: Has seen its annual AI expenditure jump from almost zero to roughly USD 25 million in just 18 months.
Despite these massive current investments, Vembu remains optimistic about the future cost structure for businesses. He suggested that as token prices continue their downward trajectory, the era of extreme spending may give way to a much more cost-effective landscape. “I don't think we will have to spend that much in the future. It drops,” he concluded.