Zhipu launches GLM-5V-Turbo: a multimodal coding foundation model

Zhipu has launched the GLM-5V-Turbo multimodal coding foundation model. The model natively integrates visual and textual capabilities, supporting the direct generation of runnable code from design drafts and screenshots, with a context window size of up to 200k. The model leads in benchmark tests for multimodal coding and GUI agents, and is deeply compatible with agent frameworks such as Claude Code and AutoClaw, enabling AI agents to possess "visual" capabilities.