Event ·Official first-hand

Claude Opus 发布，Anthropic最强旗舰模型

Claude Opus 是Anthropic最新旗舰模型，在MMLU基准测试中达到90.2%，在HumanEval代码生成测试中达到90.1%，均超越GPT-4。它支持200K token上下文窗口，在复杂推理、多语言对话和编程等任务上表现领先。该模型采用Constitutional AI训练方法，提升了安全性和可控性。

Current read

推理编程都碾压，性价比高

11 sources0° AI heatLast updated 2026/5/13 04:33:11

Evidence

11 sources

Primary sourceClaude: Blog

2026/5/13 04:33:11View source

Related source 1Anthropic: Newsroom

2026/5/12 22:33:28View source

Related source 2宝玉

2026/5/13 19:55:08View source

Related source 3向阳乔木

2026/5/14 02:56:30View source

Related source 4AI Will

2026/5/14 07:59:13View source

Related source 5lmarena.ai

2026/5/14 16:59:00View source

Related source 6Dario Amodei Blog

2026/5/12 17:58:43View source

Related source 7The Rundown AI

2026/5/13 01:11:45View source

Related source 8IT之家

2026/5/13 07:05:18View source

Related source 9arXiv: OpenAI

2026/5/13 11:12:27View source

Related source 10TestingCatalog

2026/5/13 14:36:16View 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