English(EN) The Best AI on a $20 Plan Passed Just 12% of a Real Work Exam. Which One Is Worth Paying For?

AI模型在新基准测试中难以胜任专业任务

作者 PulseAugur 编辑部 · [1 个来源] · 2026-11-14 22:12

一项新的基准测试评估了包括GPT-6 Astra、Gemini 4 Carbon和Claude Opus 5.5在内的多款AI模型在要求严苛的专业任务上的表现。结果显示，这些先进的AI系统总共仅通过了考试的12%，表明它们在现实世界的适用性和处理复杂工作场景方面的性能仍有很大的提升空间。 AI

影响当前先进的AI模型在处理复杂专业任务方面的能力有限，表明理论性能与实际应用之间存在差距。

排序理由该集群描述了AI模型在专业任务上的基准测试评估结果。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 44 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群描述了AI模型在专业任务上的基准测试评估结果。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: product, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

Medium — Claude tag TIER_1 English(EN) · ashok nair · 2026-11-14 22:12

20美元套餐中最强AI仅通过12%的真实工作考试。哪个值得付费？

<div class="medium-feed-item"><p class="medium-feed-snippet">A new benchmark put GPT-6 Astra, Gemini 4 Carbon and Claude Opus 5.5 through hours-long professional tasks. Here is what the scores mean…</p><p class="medium-feed-link"><a href="https://medium.com/@anair3107/the-…