Mistral Releases 1 Trillion-Parameter Multimodal Model Mistral Large 4, Plans to Open Weights in Three Weeks
Mistral AI released the trillion-parameter multimodal model Mistral Large 4 (nickname Le Chonk) on Tuesday, positioned as a third route between closed-source and Chinese open-source models. Currently accessible only via public guardrail endpoint, planning to open weights within three weeks after safety testing completes; the model was trained using only 4,000 NVIDIA GPUs on Mistral's own compute, with optimized use cases including cybersecurity, finance and chip design.
Return to original text
A short introduction is retained here; the full text from the source website is not reposted. For specific wording and subsequent updates, please refer to the original article.
Open AIHOT original