Meta's 168 MW Jamnagar AI Data Center Is a Statement of Intent

Meta is building a 168 MW AI data center in Jamnagar alongside Reliance, adding large-scale inference capacity in India right as Muse Spark 1.1 goes to market. The timing is not coincidence — it is the physical layer of the same story about frontier competition.

Why India, why 168 MW

Related: GPT-5 Is Here: Everything You Need to Know About OpenAI's Most Powerful Model Yet →

168 MW is a serious AI datacenter footprint, on par with mid-size hyperscale sites. Locating it in Jamnagar puts capacity next to Reliance's power infrastructure and inside the world's second-largest AI developer market. Latency for Indian users drops meaningfully, and the pricing story for Muse Spark tokens gets more credible.

Reliance as a partner

Related: Will AI Coding Agents Replace Developers? We Asked 100 Engineers →

Reliance brings power, land, and political weight that would be very difficult for a foreign hyperscaler to replicate independently. Meta gets fast execution; Reliance gets a share in the AI infrastructure buildout without absorbing the model risk. It is a template other US hyperscalers will study.

The wider capex picture

Related: The 27 Best AI Tools in 2026 (Tested for 90 Days) →

This site is part of a global build-out that also includes expanded Prineville capacity and additional US and EU sites. Meta's AI capex is now competitive with AWS and Microsoft in absolute terms, which is a structural change to how the cloud market looks.

The bottom line

Related: ChatGPT vs Claude 4: Which AI Should You Actually Pay For in 2026? →

Physical infrastructure decides who can afford aggressive frontier-model pricing. Jamnagar is one more reason to take Meta's Muse Spark price cuts seriously.