Qwen3.8 27B scores 52 on Artificial Analysis
(news.ycombinator.com)
1.
2.
Claude Sonnet 5 – benchmark results
(news.ycombinator.com)
3.
GPT-5.5 hallucinates 3x more than MIT-licensed GLM-5.2
(news.ycombinator.com)
4.
GLM 5.2 Performance Benchmarks
(news.ycombinator.com)
5.
Human scientists trounce the best AI agents on complex tasks
(feeds.nature.com)