AI 精选动态
智能评分 65
This is the strongest ARC-AGI-2 performance to date by an open-source model.
AI 推荐理由
建议重点关注其在低成本推理中的实际表现。核心解读
GLM-5.2 开源模型在 ARC-AGI-2 取得 22.8% 分数，成本 $0.25，同步创下 ARC-AGI-1 77.0% 与 $0.19，且与 GPT-5.4、GPT-5.5（低推理开销）相当。
全文
This is the strongest ARC-AGI-2 performance to date by an open-source model.
> **引用原帖 ARC Prize (@arcprize):**
> GLM-5.2 from @Zai_org on ARC-AGI (Verified)
> - ARC-AGI-2: 22.8%, $0.25
> - ARC-AGI-1: 77.0%, $0.19
> Performance is comparable with GPT-5.4 & 5.5 (Low Reasoning Effort) https://t.co/beYeeTpQJR
> https://x.com/arcprize/status/2069845152773099854