Ant Group Releases Open Source Multimodal Large Model Ling-3.0-flash-VL

On September 9, Ant Group released and open-sourced the multimodal large model Ling-3.0-flash-VL. This model is based on the MoE architecture of Ling-3.0-flash, with a total parameter count of 124 billion, activating 5.5 billion parameters for each inference. It supports image, text, and video inputs, with a context window reaching 256K tokens. Ling-3.0-flash-VL focuses on exploring how to complete real tasks more reliably and efficiently, particularly by enhancing text intelligence through the introduction of visual capabilities. The model employs a visual feedback mechanism, transforming task execution into a closed loop of observation, action, verification, and correction, thereby enhancing the reliability of execution results. At the same time, Ling-3.0-flash-VL inherits the core advantages of Ling-3.0-flash in Agent workflows, balancing output quality and execution efficiency, and advancing complete tasks at lower costs and in shorter time frames.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.