mixture-of-experts german-language sovereign-ai long-context on-premise

Kolibri delivers 1M-token context at 3B active parameters

MoE model optimized for German/English with 78B total, 3B active parameters—on-premise deployable with abstention training for regulated sectors, available on Hugging Face under Apache 2.0.

Summary

Enables sovereign deployment in compliance-heavy domains (public admin, aerospace, manufacturing) without exfiltrating data to third-party inference. On-device inference reduces latency and regulatory friction for teams building mission-critical EU/German workloads.

Why it matters

Implementation verdict

Replaces closed-model inference for regulated German-language tasks. Requires on-premise GPU capacity and familiarity with MoE routing. Worth evaluating now if you target public sector or automotive verticals—benchmarks show parity with 4x larger active-parameter models (Nemotron 3 Super 120B-A12B). Production-ready; full model weights and tech report available.

Sources

1.78B total parameters, 3B active
2.context lengths of up to 1M tokens
3.Kolibri matches models with up to four times its active parameter count, such as Nemotron 3 Super
4.21.3% of the pre-training tokens are German
5.trained to say "I don't know" when the answer isn't in the context
6.The model can be downloaded with the full weights on Hugging Face and used under the Apache 2.0 license terms

Dev Signal

Get briefs like this in your inbox — free, every weekday.

100+ sources compressed into one 4-minute read. Ranked, cited, implementation-ready.

Read the full issue →All briefs