GLM-V is an open-source multimodal vision-language model family from Z.ai (the team behind the GLM series) covering GLM-4.6V, GLM-4.5V, and GLM-4.1V-Thinking. These models achieve versatile multimodal reasoning via scalable reinforcement learning, supporting text, images, video, documents, GUI agents, and 3D spatial tasks. GLM-4.6V features a 128K context window and native multimodal tool use, accepting images directly as tool parameters.