Aleph Alpha 发布开源权重模型 Kolibri，主张欧洲 AI 主权
原标题：Aleph Alpha releases Kolibri, an open-weight model that makes the case for European AI sovereignty
AI 摘要
德国 AI 公司 Aleph Alpha 发布开源权重模型 Kolibri，这是一个 78 亿参数的德英双语模型，采用混合专家架构，每个 token 约激活 30 亿参数。公司称其在质量与运营成本上处于帕累托前沿，训练数据中德语占 21.3%，并使用中国模型生成合成数据。模型面向公共管理、航空和工业领域，依据欧盟法律并考虑 EU AI Act 开发，支持最高 100 万 token 上下文，在德国和芬兰的 768 块 B200 GPU 上训练，权重以 Apache 2.0 许可在 Hugging Face 发布。
正文节选
Aleph Alpha releases Kolibri, an open-weight model that makes the case for European AI sovereignty Aleph Alpha has released Kolibri, a German-English language model with 78 billion parameters, about three billion of which are active per token through a mixture-of-experts architecture. Aleph Alpha claims Kolibri sits on the Pareto front of quality and operating cost in both languages, outperforming compared models with similar architectures, some significantly older (March, April 2026), on either