Grok 4.5在单智能体场景下比Opus还强，又快又聪明，编程和对话都顺手。
Grok 4.5由SpaceXAI与Cursor合作训练，是首个在单智能体迭代任务中表现优于Claude Opus的模型。用户测试反馈其速度快、反应直接且智能度高，能轻松绕过编程障碍。该模型是Grok系列首次面向软件工程之外通用场景构建。
This is the first model that I personally prefer over Opus when I am iterating with a single agent. ...
This is the first model that I personally prefer over Opus when I am iterating with a single agent. Fast Grok 4.5 feels very very good! The model always finds its way around seemingly annoying roadblocks, is very direct and easy to communicate with, and is just really really intelligent. Really excited about our future work with @SpaceXAI Cursor @cursor_ai We've partnered with SpaceXAI to train Grok 4.5. It’s our most powerful model yet and the first we've built for more than software engineering. 🔗 View Quoted Tweet 💬 10 🔄 4 ❤️ 155 👀 5216 📊 22 ⚡