Event ·Multi-source

Claude Fable 5.1 发布

Claude Fable 5.1 在 Terminal-Bench-Science 0.1 基准测试中得分为 52.6%，较 Fable 5 的 24.7% 有显著提升。该模型提供五种推理级别：低、中、高、超高和最高。在生成 SVG 鸬鹚图像任务中，最高推理级别产生了最佳结果，消耗 65,927 个输出 tokens，耗时 13 分 54 秒，成本 3.30 美元。

Current read

Anthropic 发布了 Fable 5.1，在科学基准测试中大幅领先，推理级别设置对创意输出质量影响显著。

8 sources73° AI heatLast updated 2026/9/1 23:57:28

Evidence

8 sources

Primary sourceSimon Willison’s Weblog

2026/9/1 23:57:28View source

Related source 1Claude

2026/9/1 18:03:16View source

Related source 2宝玉

2026/9/1 18:27:16View source

Related source 3shao__meng

2026/9/1 23:39:07View source

Related source 4Decoder

2026/9/1 20:01:55View source

Related source 5marktechpost

2026/9/1 20:30:21View source

Related source 6IT之家

2026/9/1 22:28:51View source

Related source 7Stanford AI Lab

2026/9/3 21:56:12View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story