1010B / 68.8B active (MoE)
Apache 2.0
Trillion-scale multimodal foundation model for enterprise scenarios; unified architecture combining vision encoder, language backbone, and multimodal alignment module; MoE language backbone with 103 Transformer layers; pretrained from scratch at 1515B parameters and reduced to 1010B via Layer-Adaptive Expert Pruning (LAEP), reporting 49% pretraining-efficiency improvement; activates 68.8B parameters per token; incorporates Localized Filtering-based Attention (LFA) and upgraded Reflection Inhibition Reward Mechanism (RIRM); official README reports strong enterprise-oriented benchmark performance across RAG, document/table understanding, coding, and multimodal tasks.