Event ·Multi-source

Grok 4.7 上线 Devin，FrontierCode 1.1 Extended 得分 59.4%

Cognition 宣布其编程智能体 Devin 正式接入 xAI 的 Grok 4.7 模型。在 FrontierCode 1.1 基准的 Extended 难度上，Grok 4.7 拿到 59.4% 的分数。Cognition 在自己的评估中发现，该模型在难度较高的后端工程任务上表现很好。

Current read

Devin 现在能调用 Grok 4.7 了，Cognition 实测它在难的后端工程任务上表现好，写后端的可以试试。

11 sources78° AI heatLast updated 2026/9/21 21:27:53

Evidence

11 sources

Primary sourceCognition

2026/9/21 21:27:53View source

Related source 1Windsurf

2026/9/21 21:28:54View source

Related source 2Genspark

2026/9/23 02:45:22View source

Related source 3OpenRouter

2026/9/21 16:31:16View source

Related source 4lmarena.ai

2026/9/21 16:37:46View source

Related source 5Decoder

2026/9/21 16:56:36View source

Related source 6eric zakariasson

2026/9/21 22:01:00View source

Related source 7Guillermo Rauch

2026/9/22 22:03:10View source

Related source 8shao__meng

2026/9/23 01:44:23View source

Related source 9宝玉

2026/9/21 17:34:10View source

Related source 10AI Will

2026/9/22 01:18:18View 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