智能体
x.com ·
@Arena.ai (@arena)
GPT-6 Astra在Agent Arena中排名第二，成本效益显著
Arena.ai公布GPT-6 Astra (Max)在Agent Arena的评测结果，以+12.5%的净改进排名第二，用户好评率提升40.6%，确认成功率达21.9%，每次任务成本4.01美元，重塑了性能-成本平衡。
AI推荐理由 Arena.ai发布GPT-6 Astra在Agent Arena的详细评测数据，包括改进幅度、用户反馈和成本，信息具体且来源可信，对Agent性能评估有重要参考。
85
浏览 46,099
点赞 439
评论 15
转发 27
正文
Agent Arena results for GPT-6 Astra (Max) by [@OpenAI](https://x.com/OpenAI) are in: it debuts at #2 overall with +12.5% net improvement across +8.9K real-world agentic sessions!
GPT-6 Astra (Max) scored a massive +40.6% in Praise vs. Complaint (implicit sentiment in user reactions), and +21.9% in Confirmed Success (explicit user feedback that the task worked).
With a $4.01 cost per median task, it also reshapes the Pareto frontier for Agent Arena. More info on Astra’s performance compared to cost below.
Congrats to [@OpenAI](https://x.com/OpenAI) on this release!
OpenAI@OpenAI
This is GPT-6 Astra.
Anything you can do on a computer, Astra can do for you. Fast.
[Your browser does not support the video tag.](https://video.twimg.com/amplify_video/2095595661559574528/vid/avc1/1920x1080/grXdGA07X-yPH_8z.mp4?tag=29)
[🔗 View on Twitter](https://x.com/OpenAI/status/2095595741528125780/video/1)
[🔗 View Quoted Tweet](https://x.com/OpenAI/status/2095595741528125780)
💬29🔄43❤️655👀72841📊83