Mistral AI has launched Mistral Large 4, a new open-weights, multimodal model available in public preview. This model boasts 1.05 trillion parameters, utilizing a Mixture-of-Experts (MoE) architecture where only 49 billion parameters are active per token. It includes a 1.6 billion parameter visual encoder for image processing and supports a context window of one million tokens. AI
IMPACT This release expands multimodal capabilities and context window sizes, potentially influencing enterprise adoption and competitive pressures in the LLM market.
RANK_REASON Frontier-lab model release with system card. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
Read on Mastodon — mastodon.social →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →