GLM-4 & GLM-4V
Zhipu AI's bilingual powerhouses with 128k context and multimodal vision
Models
In plain English
A versatile open model family from Zhipu AI that excels at coding, complex tool use, and understanding images right on your desk.
Technically
Open foundation models from Zhipu AI (e.g. GLM-4-9B, GLM-4V-9B) utilizing 2D RoPE positional encodings for multi-resolution vision and 128k text contexts. Features robust tool-calling compliance and bilingual English/Chinese benchmark performance.