Z.ai Chief Scientist Reveals GLM-5.5 Training Pipeline and Sizes
Z.ai chief scientist Jie Tang has revealed details regarding the training pipeline and upcoming model sizes for the forthcoming GLM-5.5 series. The new family of models will include 32B, 122B, and 256B configurations. This announcement offers the local LLM and open-weights community a valuable early look at upcoming high-performance AI architectures. The introduction of a 32B variant is especially exciting for developers running models locally on consumer hardware. The GLM-5.5 lineup features a diverse range of scales, specifically 32B, 122B, and 256B parameters, catering to different deployment needs. Specific performance benchmarks and official release dates have yet to be fully detailed.
## BACKGROUND
Zhipu AI (Z.ai) is a prominent Chinese AI startup and research lab originating from Tsinghua University that develops the GLM series of large language models. Jie Tang is a chair professor at Tsinghua University and the co-founder and chief scientist of Zhipu AI.