UK sovereign LLM inference | Hacker News

Civo’s relaxAI has launched a UK-hosted, OpenAI-compatible inference API built on open-source models and NVIDIA Blackwell GPUs.

The service promises lower token costs while keeping data in the UK, which could appeal to regulated sectors and privacy-conscious buyers.

The launch has sparked debate over what “sovereign AI” really means, especially when foreign models and chips are still part of the stack.

It also highlights growing demand for local AI infrastructure shaped by cost, compliance, and concerns about overseas control.