Skip to content

Back to the catalog

Sahil Verma (@sahilvermaai) on X
x.comthread

Sahil Verma (@sahilvermaai) on X

Aleph Alpha released Kolibri, a 78B-parameter open-weight mixture-of-experts model with only 3.46B active parameters per token, a 1M token context window, and Apache 2.0 licensing with full weights on Hugging Face. The model was trained on German and Finnish infrastructure, positioning it as a European-built alternative to US/Chinese open-weight releases.

TL;DR Aleph Alpha's Kolibri: a 78B MoE model (3.46B active) with 1M context, Apache 2.0, trained in Europe.
Open original ↗

Organisations named here

  • Aleph Alpha
  • Hugging Face

Technologies named here

  • Kolibri
  • Hugging Face
  • Mixture of Experts

Techniques named here

  • Mixture of experts

Related in models