SiliconFlow· @SiliconFlowAI · X·· 2026-06-23AI 评分59
AI 导读
硅基流动测试了 GLM-5.2、GPT-5.5、Opus 4.8 和 GLM-5.1 的相同提示。据 @arena 引用，GLM-5.2 (Max) 在 Code Arena: Frontend 排名第 2，以 +29 分领先 Claude Opus 4.7 (Thinking)，仅次于 Fable 5；是最好的开源模型，大幅超越 Kimi-K2.6 和 Minimax-M3，并在 React（第 2）、HTML（第 4）及品牌营销、参考设计、数据分析等多个子类别中位居第一。主推文指出，在 SiliconFlow 上使用 GLM-5.2 可获得 Opus 级前端生成能力，输入成本降低约 3.6 倍，输出成本降低约 5.7 倍。
正文 · 原文
What happens when frontier models face the same prompt?
We tested GLM-5.2, GPT-5.5, Opus 4.8, and GLM-5.1.
And the result: GLM-5.2 closed the performance gap with Opus 4.8 at the cost of friction.
Get Opus-level frontend generation with GLM-5.2 on SiliconFlow—at ~3.6× lower input cost and ~5.7× lower output cost
Let's build more & spend less today😈
https://cloud.siliconflow.com/models?target=zai-org/GLM-5.2
Exciting news: GLM-5.2 (Max) ranks #2 in Code Arena: Frontend, with +29pt over Claude Opus 4.7 (Thinking) and only behind Fable 5! GLM-5.2 is the best open model vs Kimi-K2.6 and Minimax-M3 by a large margin. - #2 React and #4 HTML sub-leaderboards - Ranks as the top model in nearly all sub categories: Brand & Marketing, Reference-Based Design, Data & Analytics, Consumer Product, Gaming, and Simulations. Congrats @Zai_org for the incredible milestone!在 X 查看被引用的帖子
来源：SiliconFlow · x.com