Nobodywho

Nobodywho is an inference engine that runs AI models directly on devices and on company servers, with no calls to an external model provider. It is built and maintained in Copenhagen and released under the EUPL 1.2, the European Union's open source licence, with development done in public on GitHub. The engine ships as open source SDKs for Rust, Python, Kotlin, Swift, Flutter, React Native and Godot. That makes it usable inside phones, laptops, embedded hardware, games and other apps, where it runs compact open models fully offline. For organisation wide use, it can instead run larger open models on owned or leased GPU servers inside the organisation's own network. The two models can be combined, with work orchestrated across devices and servers. On top of the engine, the vendor sells a managed service called the Assurance Layer. It covers installation and configuration, model selection tuned to each use case, support when issues arise, assessment and rollout of newer open models, governance controls that restrict which models may be used, and bespoke tooling shaped around existing workflows. It is aimed at enterprises that want local AI without an in house AI team. The engine carries no per token costs and keeps running without the vendor's service or an internet connection. The site publishes no prices for the Assurance Layer; interested organisations are directed to contact the team.