Anthropic 内部最强模型“Model 2”仅限内部使用

AI 摘要

Anthropic 正在内部使用一款代号为“Model 2”的未发布 AI 模型，其综合性能略强于公开的 Claude Mythos 5，但在某些领域较弱。该模型在内部能力指数上比 Mythos 5 高约 1.5 点，主要用于编码、数据生成和研究工程。目前公司没有对外发布该模型的计划，并评估其错位风险为低。

正文节选

Anthropic's most capable model, codenamed "Model 2," is for internal use only Anthropic is running an unreleased AI model internally that outperforms every publicly available version of Claude. That's according to the company's Risk Report from August 2026. The report calls the model "Model 2" and places it in the Mythos class. Anthropic says it's slightly stronger overall than Claude Mythos 5, but weaker in some areas. It doesn't show a big capability jump like the one from Opus 4.6 to Mythos.

发布时间：2026-08-20 18:04

抓取时间：2026-08-20 19:07

来源机构：THE DECODER