Thread by @SwarajyaMag

In 2021, Mukesh Ambani announced a 40 GWh cell gigafactory at Jamnagar, scaling to 100.

In May 2026, Reliance is back in talks with CATL — to buy finished cells and package them into Indian boxes, with press releases still calling it manufacturing. 🧵

Reliance is not alone. Exide's cells are licensed from SVOLT, Amara Raja's from Gotion and Highstar, Tata's Agratas from AESC — the Chinese-owned maker that has a board seat.

Every conglomerate that set out between 2021 and 2023 to build cells has retreated to assembling them.

Licensing now, R&D later, has been a plausible playbook — CATL was a TDK licensee, Samsung SDI started on Japanese chemistry.

But the trick is to pair the licensing with R&D investment at scale, and the Indian conglomerates have done the licensing without the R&D.

CATL spent $2.58 billion on battery R&D in 2024 and the Korean Big Three $2.1 billion combined, while Reliance's entire group R&D budget across all sectors that year was $437 million. CATL alone spent six times that on batteries.

A turnkey BESS sold globally for $117 per kilowatt-hour in 2025, of which $75 was core equipment from China, mostly the cells.

Indian assemblers capture 40% of the system value; rest stays in China, with cell margins holding while integration margins compress.

On the current trajectory, the Indian battery industry will be born already dependent.

The capital and the talent are in place. What is missing is the corporate appetite for the long, expensive half of the playbook.

Full piece in @SwarajyaMag.

swarajyamag.com/tech/from-cell…