vivo launches four blue mind large models with enhanced capabilities

On September 16, 2026, vivo unveiled an upgraded Blue Mind large model during its Developer Conference, introducing a matrix of four core models: Blue LM-RealTime, Blue LM-Nano, Blue LM-Flash, and Blue LM-Pro. These models are designed to enhance speech understanding, edge-side memory, and cloud-side task execution. Blue LM-RealTime focuses on speech recognition, offering robust performance in noisy environments and handling accents, continuous speech, and mixed languages. Blue LM-Nano, an edge-side model, excels in real-time perception and memory, integrating data from screens, images, videos, and text while tracking user habits and device status. Blue LM-Flash and Blue LM-Pro, cloud-side models, leverage an Agentic engine and system-level collaboration to execute complex tasks across devices. Users can describe needs, and the system will act as a task controller, adapting to individual preferences through continuous learning. vivo emphasized that the edge-side models maintain industry-leading capabilities, achieving multiple top rankings in performance metrics.