Aleph Alpha Launches Sovereign AI Model Kolibri
Aleph Alpha released Kolibri, a sovereign AI model for mission-critical government and industrial applications.
European AI companies are increasingly competing on sovereignty, compliance and domain fit rather than raw model scale alone.
What happened
Germany’s Aleph Alpha launched Kolibri, a specialised language model for mission-critical government and industrial use. The model has 78B total parameters, around 3B active per token, and supports retrieval-augmented generation, tool calling and agentic workflows.
Kolibri is optimised for German-language applications and positioned for environments where data control, reliability and compliance matter.
Why it matters
Many public-sector and industrial customers cannot treat AI as a generic cloud API. They need systems that handle local language, regulatory requirements, deployment constraints and sensitive data.
Kolibri is a useful signal for the European sovereign AI market: smaller active models, controlled deployment and domain-specific reliability may be more important than racing US frontier labs model-for-model.
The bigger picture
AI sovereignty is moving from political slogan to product strategy. European model companies are carving out demand in government, industry and regulated sectors where control and trust can outweigh pure benchmark leadership.