Aleph Alpha Open-Sources Kolibri: 78.1B-Parameter English-German MoE Activating Only 3.46B per Token

Original sourceMarkTechPost

Summary

Aleph Alpha released Kolibri, an open-weight MoE language model with 78.1B total parameters focused on English and German, activating only 3.46B parameters per token, about 4.4% utilization. Features include up to 1,048,576-token context, adjustable inference effort on demand, and Apache 2.0 licens…

Key points

For teams that need to deploy bilingual models in their own or regulated environments, this is a concrete option they can download and operate themselves.
An efficient open-source MoE with low activated parameters lets sovereignty and compliance use cases deploy long-context bilingual capabilities at lower hardware thresholds.
Developers can deploy a million-token-context model on one or two GPUs, lowering self-hosted inference costs, but must bear operational and compliance responsibility themselves.

開源模型

Editorial note

This page is Code & Chain's editorial summary of public sources. It may be prepared with AI assistance and published through an automated workflow. Refer to the original sources; this content is not investment, legal, or tax advice.