Colibrì Runs GLM-5.2 on 16GB of RAM by Offloading to SSD
Open-source inference engine Colibrì streams MoE model weights from an NVMe SSD, letting a 16GB-RAM machine run Zhipu AI's 744-billion-parameter GLM-5.2.
GLM News and Analysis
10 verified stories covering GLM, product updates and industry developments.
Zhipu's GLM-5.3-FlashX Is 5x Faster, 2.5x Pricier
Zhipu's GLM-5.3-FlashX hits 200 tokens/s, 5x faster than Flash, but costs 2.5x more — betting concurrency headroom can be priced as a premium.
DeepSeek Retakes OpenCode Top Spot as GLM Goes Paid
DeepSeek retook OpenCode's daily usage lead the moment GLM-5.3-Flash started charging, ending its 41-trillion-token free run.
Zhipu Claims Ox Alpha, Usage Tops DeepSeek by 2x
Zhipu confirms Ox Alpha is its new GLM model after anonymous OpenRouter testing, with free-preview usage already more than double DeepSeek's.
GLM-5.2 closes in on closed coding models
Z.ai’s open-weight GLM-5.2 scored 74.4 on FrontierSWE, near Claude Opus and above GPT-5.5 in some coding benchmarks, while keeping token prices far lower.
Zhipu opens GLM-5.2 with a 1M-token context
Zhipu's GLM-5.2 pairs a 744B-parameter MoE design with a 1M-token window and a coming MIT-style open-weight release for coding agents.
Zhipu AI Launches GLM-5, Reportedly Plans Hong Kong IPO
Zhipu AI has released its next-generation GLM-5 model and is reportedly preparing for an initial public offering on the Hong Kong Stock Exchange, which would make it one of the first major Chinese AI model companies to go public.
Zhipu Brings GLM-4.6 to Chinese AI Chips
Zhipu is adapting GLM-4.6 for multiple domestic AI chip platforms as export controls make hardware resilience a strategic priority.
Zhipu GLM-4.7 Focuses on Production Stability
Zhipu AI's latest GLM-4.7 model prioritizes consistency and reliability in production environments over benchmark scores, targeting enterprise customers with improved output stability, long-running agent support, and Chinese-language optimization.
Zhipu GLM-4.5 Runs Full Model on Just 8 H20 GPUs
Zhipu AI optimized GLM-4.5 for NVIDIA's China-specific H20 GPU, enabling the full model to run on just eight cards, significantly lowering deployment costs for domestic enterprises.