Zhipu
Chinese LLMs top Claude in Civilization V — turn 100 is the real test
CivBench stress-tests LLMs with full Civilization V games. Zhipu's GLM-5.3 beats Opus-5.5; Alibaba's Qwen-3.8-27B surprises. Single-turn Q&A misses th
Back to home
Alibaba Qwen
3 articles tagged with this topic
H Company
Alibaba's Qwen Becomes Base for Europe's Open-Source Computer-Use AI, Challenging Anthropic
H Company launched open-source Holo4 (27B/35B-A3B): AI views screenshots to click, type, code. Built on Alibaba Qwen, challenging Anthropic.
Zhipu AI
GLM, Qwen Catch Closed-Source Leaders on Agent Benchmarks in Just Two Months
GLM and Qwen now match top closed-source models on Agent Arena Code — a result unthinkable two months ago.