Aleph Alpha releases Kolibri, an open-weight English-German mixture-of-experts language model on Hugging Face
Open source release Provisional 86% confidence first seen
Aleph Alpha published its Kolibri model as an open-weight mixture-of-experts language model on Hugging Face, positioning it as a “sovereign” alternative for regulated deployments. Coverage describes a 78B-parameter model with a smaller active parameter count per token and a context window of about 1 million tokens, alongside details on licensing and deployment options.