Aleph Alpha has released Kolibri-1, a mixture-of-experts reasoning model designed for German and English languages. This model is optimized for long-context processing, supporting up to 1,048,576 tokens, and features an explicit reasoning mode and tool-calling capabilities. Kolibri-1 aims to provide strong performance at a low serving cost by activating only a fraction of its parameters per token, making it suitable for human-AI collaboration in applications like conversational assistants and document processing systems. AI
IMPACT Enhances capabilities for German and English language tasks, particularly in long-context reasoning and agentic workflows.
RANK_REASON Frontier-lab model release with system card. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
Read on Hugging Face Trending Models →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →