Awesome-MCoT

Why we included this project

Multimodal chain-of-thought reasoning is moving fast, and this repository is the companion index to the first systematic survey of the field. It is a paper-by-paper map that tracks step-by-step reasoning in multimodal large language models across images, video, speech, audio, 3D data, and structured inputs, paired with a taxonomy and a running timeline from the arXiv paper. That makes it easy to see how techniques like MCTS, instruction tuning, and slow-thinking models have evolved, and which applications, from robotics to healthcare to autonomous driving, are getting real attention. For teams weighing whether to build on reasoning-capable multimodal models, the list is a practical way to find the key papers worth reading before committing to an approach.