NobodyWho raises €2 million to run language models locally, preserving privacy and cutting carbon

App developers face surging costs and privacy concerns related to cloud-based large language models. These models require constant server access, transmit user data to remote providers and consume significant energy, all while facing increasing scrutiny under the GDPR and mounting demands for greater sustainability.

Copenhagen-based company NobodyWho offers an open-source inference engine that powers small language models on everyday devices such as laptops and phones. Using Vulkan, Metal, and SIMD, it enables rapid local processing without the need for an internet connection or data sharing. The company, founded by Cecilie Waagner Falkenstrøm and Asbjørn Olling, enables seamless integration via simple code snippets and supports thousands of GGUF models for games, apps, and NPCs. It also slashes carbon footprints by up to 500x through edge computing.

NobodyWho has secured €2 million in pre-seed funding from PSV Tech, The Footprint Firm, and Norrsken Evolve to continue developing its technology.

Sources: NobodyWho | Crunchbase

Founders: Cecilie Waagner Falkenstrøm, Asbjørn Olling