Event ·Multi-source

GPT-6 Astra 和 Claude Fable 5.1 性能提升但成本更高

我们对比了不同模型家族的最佳表现。GPT-6 Astra 的性能提升了 11.7%，但每任务成本为 3.94 美元。Claude Fable 5.1 的性能提升了 13.7%，但每任务成本为 4.40 美元。GPT-5.6 Sol 的性能提升了 7.0%，每任务成本为 1.03 美元。Claude Opus 5 的性能提升了 10.2%，每任务成本为 2.07 美元。

Current read

OpenAI 和 Anthropic 新发布的模型性能提升明显，但价格也更高，对比一下不同模型的性价比。

11 sources58° AI heatLast updated 2026/9/16 19:13:56

Evidence

11 sources

Primary sourcelmarena.ai

2026/9/16 19:13:56View source

Related source 1DeepLearning.AI

2026/9/15 14:30:37View source

Related source 2IT之家

2026/9/15 09:42:25View source

Related source 3掘金本周最热

2026/9/17 04:51:39View source

Related source 4theverge

2026/9/18 15:30:16View source

Related source 5Decoder

2026/9/18 17:20:30View source

Related source 6a16z

2026/9/14 22:15:16View source

Related source 7e27

2026/9/15 02:30:13View source

Related source 8Greg Brockman

2026/9/15 18:29:47View source

Related source 9AI Will

2026/9/16 01:17:08View source

Related source 10Gary Marcus

2026/9/16 04:22:08View 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