Zhipu releases the native multimodal coding foundation model GLM-5V-Turbo, deeply adapted for OpenClaw

Sina Tech News, April 2nd — Zhipu released its first native multimodal Coding base model, GLM-5V-Turbo, announcing that the model deeply integrates visual and programming capabilities, capable of natively processing multimodal information such as text, images, and videos, while also excelling in programming, long-term planning, operation execution, and other complex tasks.

It is reported that GLM-5V-Turbo has achieved leading performance in core benchmarks such as multimodal Coding and Agent with a smaller size, introducing visual capabilities while maintaining the same level of pure text programming and reasoning abilities. Additionally, it is deeply adapted to Claude Code and Lobster scenarios, enabling OpenClaw Lobster to have true visual capabilities and understand on-screen information.

Unlike traditional pure text Coding models, GLM-5V-Turbo can directly interpret visual information such as design drafts, webpage screenshots, and K-line charts, and generate executable code, achieving a “what you see is what you get” AI programming experience. Currently, the model has been made accessible through Zhipu’s MaaS platform. (Wen Meng)

Massive information, precise interpretation, all on Sina Finance APP

Editor: Yang Ci