Ling-V2

Ling-V2 is an open-source family of Mixture-of-Experts (MoE) large language models developed by InclusionAI. It introduces highly sparse architectures, enabling models to achieve strong reasoning and instruction-following capabilities while remaining computationally efficient.