Locai

Locai is a London-based deep-tech startup building infrastructure that moves AI model inference from the cloud onto end-user devices. Its OpenAI-compatible endpoint is designed as a drop-in replacement for cloud APIs, letting SaaS and desktop AI vendors migrate existing code with a single line change while compute runs on customers' own laptops, workstations, or dedicated hardware.

The product targets companies whose unit economics are eroded by per-API-call cloud inference billing — meeting tools, writing platforms, code assistants, customer support apps — and those with strict data-sovereignty constraints. By processing prompts locally, Locai removes the variable cost tied to user growth, eliminates third-party data routing, and supports air-gapped deployments. The technical premise rests on recent advances in consumer and enterprise hardware that allow modern devices to run 7–13 billion parameter models locally.

Locai was founded by Joseph Ward and Saif Al-Ibadi, who previously built defence and aerospace deep-tech company SmallSpark Space Systems. The team participated in the initial cohort of the Google for Startups Accelerator in 2025 and has demonstrated its approach through a multi-year contract with B2Space, deploying AI agents at the edge of space with a reported 90%-plus reduction in bandwidth costs. The company has now raised £1 million in pre-seed funding led by Fuel Ventures.