SemiAnalysis releases GDPx benchmark for knowledge work; no AI agent scores above 12%
What happened
The Decoder reported on November 6 that SemiAnalysis released GDPx, a benchmark simulating knowledge work across 11 domains and 206 tasks. No AI agent scored above 12%. Claude Opus 5.5 led at 12%, while Gemini 4 Carbon scored 10.9% at $16.29 per task, about a third of Claude Opus 5.5's cost. Excluding cybersecurity, biology and alignment, GPT-6 Astra topped the field at 11.8%.
Written by AI from the coverage · updated 6 hours ago
Coverage
Follow the reports to see the story from different sides.
- The DecoderNo AI agent scores above 12 percent on SemiAnalysis' new real-world work benchmark
SemiAnalysis 发布模拟知识工作的 GDPx 基准，覆盖 11 个领域的 206 项任务，没有 AI 智能体得分超过 12%。Claude Opus 5.5 以 12% 领先，Gemini 4 Carbon 为 10.9%，每项任务成本为 $16.29，约为前者的三分之一；剔除网络安全、生物学和对齐领域后，GPT-6 Astra 以 11.8% 居首。
Heat over time
Heat now 8·Comparable peak 10（Nov 6 02:00）·Comparable change over 24 hours –
The trend only compares accounts observed without gaps, so its range may be smaller than the current heat. Hover or tap the chart for each hour; the left and right arrow keys step through it.