A new AI-driven paradigm for LIBs’ innovation. We propose an industry-level pretrained AI model tailored for lithium-ion batteries, battery large model, driving intelligent upgrades across research and development (R&D), manufacturing, and lifecycle management. Foundation layer: integrates material, cell, manufacturing, process and operation and maintenance (O&M) databases, constructing a comprehensive industry-specific knowledge graph. Intermediate layer: employs time-series models, hybrid AI-physics models, multimodal AI models (text, images, etc.) and reinforcement learning (RL) models to enhance cell design, lifespan prediction, and process optimization. Application layer: leverages AI-generated content (AIGC) to autonomously generate design solutions, predict battery performance, and detect defects, enabling full-lifecycle optimization. This AI-driven framework transforms lithium battery development, paving the way for data-driven innovation, autonomous optimization, and intelligent manufacturing.