Skip to content
TopicTracker
From HackerNewsView original
TranslationTranslation

GLM-5.2 is the new leading open weights model on Artificial Analysis

GLM-5.2 has become the top-performing open weights model on the Artificial Analysis Intelligence Index, surpassing previous leaders in benchmark evaluations.

Background

- **Artificial Analysis** is a third-party benchmarking site that tests LLMs on quality, speed, context length, and price. Its "Intelligence Index" is a widely followed leaderboard. - **GLM-5.2** is the latest model from **Zhipu AI**, a Beijing-based company. It uses a **Mixture of Experts** architecture — a huge total parameter count but only a fraction activates per query, balancing capability with cost. - GLM-5.2 is an "open weights" model (publicly released trained parameters, not fully open-source), and it now **tops the Intelligence Index** among open-weights models, surpassing Meta's Llama 4. - This signals that Chinese open-weight models, which have been steadily improving, are now directly competitive with — or ahead of — Western ones on independent benchmarks.

Related stories

  • The author tests the new open-weights GLM 5.2 model against Gemini 3 Flash by having them play text adventure games within a fixed budget of $0.15 per attempt. A harness tracks achievements earned in each game to compare the models' performance.