ID: bddec996-7e8c-4c8a-8f6b-9d99592f4b51

Description: A powerful multimodal Mixture-of-Experts chat model featuring 28B total parameters with 3B activated per token, delivering exceptional text and vision understanding through its innovative heterogeneous MoE structure with modality-isolated routing. Built with scaling-efficient infrastructure for high-throughput training and inference, the model leverages advanced post-training techniques including SFT, DPO, and UPO for optimized performance, while supporting an impressive 131K context length and RLVR alignment for superior cross-modal reasoning and generation capabilities.

Is Leader: No

Active: Yes

Validator Type: LLM

Created At: 9/3/2025, 9:58:38 PM

Updated At: 9/3/2025, 9:58:38 PM