Local-first tooling for working safely with LLMs and autonomous AI — currently focused on Dutch.
We build small models and companion tools that keep data on your machine. Privacy, grounding, governance — shipped as open source. The architecture is language-agnostic; today's models and rules target Dutch, with other languages planned.
- Filenthropist — local-first file scanner and labeler that lets you work safely with autonomous AI agents. Multilingual by design, currently focused on Dutch PII and GDPR.
pipx install "filenthropist[all]"
Our models live on Hugging Face: huggingface.co/LokaalHub.
nl-lokaal-middel— Dutch PII NER, 473 MB, F1 0.84nl-lokaal-klein— Dutch PII NER (fast), 181 MB, F1 0.78- Next:
nl-nli-*(Dutch NLI / groundedness scoring)
Models follow <task-in-local-language>-<size-tier>:
klein ≤ 200 MB (throughput), middel ≤ 500 MB (accuracy), groot reserved.
SMEs and public institutions operating under GDPR and the EU AI Act often can't send documents or queries to foreign cloud APIs for compliance or procurement reasons. We build tools and models compact enough to run on a laptop so local deployment is a first-class option, not a compromise. We're starting with Dutch because that's the market we know; the architecture generalises to other languages.
Questions or collaboration? Open an issue on any of our repos.