Anthropic发布了Claude Sonnet 5，这是一款新的中端AI模型，展示了更强的代理能力。最新版本在所有基准测试中均超越了其前代产品Sonnet 4.6，在代理编码测试（特别是SWE-bench Pro）中取得了63.2%的得分。该公司还强调，Sonnet 5比旗舰模型Opus 4.8更具成本效益，在8月31日之前，API的初始定价为每百万输入令牌2美元。 AI
影响 在代理编码基准测试中创下新的SOTA（State-of-the-Art）；给竞争对手带来匹配性能和成本的压力。
排序理由 前沿实验室模型发布，带有系统卡[lever_c从frontier_release降级：ic=2 ai=1.0]
在 Mastodon — fosstodon.org 阅读 →
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
Anthropic has released Claude Sonnet 5, its most agentic mid-tier model yet. The new model beats Sonnet 4.6 on every benchmark, scoring 63.2% on agentic coding tests, while costing less than the flagship Opus 4.8 at 2 USD per million input tokens. https://www. marktechpost.com/20…
Anthropic has released Claude Sonnet 5, its most agentic mid-tier model yet. The new model beats its predecessor on every benchmark, scoring 63.2% on SWE-bench Pro for agentic coding, while costing less than the flagship Opus 4.8. Intro pricing is 2 USD per million input tokens t…