# OpenAI 与 Anthropic 战略分化，Astra 或超越 Claude

- 来源：Yuchen Jin (@Yuchenj_UW)
- 发布时间：2026-08-08 07:01
- AIHOT 分数：54
- AIHOT 链接：https://aihot.virxact.com/items/cmsjkcybe0c8hroo5oaedbz70
- 原文链接：https://x.com/Yuchenj_UW/status/2085863962671976518

## AI 摘要

OpenAI 与 Anthropic 在 AI 安全策略上分歧明显：OpenAI 倾向于尽快向公众发布具备网络能力的尖端模型，而 Anthropic 更为谨慎。OpenAI CEO Sam Altman 表示，Astra 模型功能强大，正努力使其广泛可用，但因网络能力需更多时间确保安全。这意味着 Astra 发布时可能成为全球最强模型，GPT 将首次领先 Claude。

## 正文

OpenAI and Anthropic are taking very different strategies.

OpenAI seems willing to ship frontier models with cyber capabilities to the public asap, Anthropic is more cautious.

That means Astra could very well be the best model in the world when it launches.

For the first time in a long time, GPT will be ahead of Claude.

### 引用推文

> Sam Altman：astra is a powerful model and we are working to make it generally available. we do not think it is a good strategy to keep powerful models to a chosen few. give...