Fractal Analytics is set to revolutionize AI in India with its proposal for the country’s first large reasoning model, backed by significant investment and government support.

Fractal Analytics is making strides in the artificial intelligence sector by proposing to build India’s first large reasoning model (LRM), with an investment of INR 118 crore. This initiative aims to enhance India’s AI capabilities and compete globally, especially against established players like OpenAI and DeepSeek.

Fractal’s Large Reasoning Model Initiative:

Fractal Analytics, a Mumbai-based AI startup, has announced plans to develop India’s first large reasoning model (LRM) as part of its broader strategy to capitalize on the growing demand for advanced AI technologies. The project entails a total investment of INR 118 crore, with INR 42 crore sourced from Fractal’s funds and an additional INR 76 crore sought from the Indian government under India’s AI Mission.

Key Features of the LRM:

• Model Variants:

The proposed LRM will include three versions: small (2-7 billion parameters), medium (20-32 billion parameters), and large (70 billion parameters) models. These models will be trained on datasets containing up to one trillion tokens, focusing on enhancing reasoning and analytical capabilities.

• Technology Investment:

A significant portion of the funding will be allocated for acquiring high-performance graphic processing units (GPUs), essential for effectively training large language models. The company anticipates needing around 700,000 hours of GPU time over nine months to train these models.

• Competitive Landscape:

Fractal’s initiative comes at a time when reasoning models are gaining traction globally. Unlike traditional conversational AI, which relies on mimicking patterns in data, LRMs are designed for complex problem-solving and decision-making tasks. This shift towards reasoning-focused models is crucial as industries increasingly seek AI solutions that can perform advanced analytical functions.

Strategic Importance for India:

The development of LRMs is seen as a pivotal step for India in the global AI race, particularly against leaders like the US and China. By focusing on post-training models that can “think and reason,” Fractal aims to position India as a significant player in the field of artificial general intelligence (AGI).India’s AI Mission has already attracted considerable interest, receiving over 187 applications from startups looking to develop foundational AI models. This reflects a growing ecosystem that supports innovation in AI technology across various sectors.

Fractal Analytics’ proposal for India’s first large reasoning model signifies a major advancement in the country’s AI capabilities. By investing in advanced reasoning technologies, Fractal not only aims to enhance its competitive edge but also contributes to India’s ambition of becoming a leader in artificial intelligence.