Ant Group releases Ling-3.2-flash multimodal model

Ant Group officially released the Ling-3.2-flash model on November 5, marking its entry into the high-capacity multimodal AI market. The model architecture utilizes 560 billion total parameters, with 25 billion parameters active during each inference step to manage complex data inputs.

The model is engineered to handle advanced visual interpretation, multi-step agent workflows, 3D interaction, and specialized applications within the medical field. By supporting a one-million-token context window, it aims to process long-form documentation and extended visual media more efficiently than previous iterations.

Ant Group has made the model available via a web-based chat interface and APIs that utilize standard OpenAI and Anthropic protocols to facilitate integration for developers. The company is offering a free trial period for the first two weeks following the release to encourage platform adoption.