Today for AI
HOT RADAR
agentHEAT 16.2°

AI CURATED HOT STORY · 11/2/2026

SemiAnalysis Launches GDPx Benchmark Showing Low AI Agent Scores on Expert Tasks

1 reports archived1 independent sourcesupdated 11/2/2026, 20:45:11
Synthesis & Latest Updates
1 Sources Cross-Validated

SemiAnalysis released the GDPx benchmark to evaluate AI agents on expert-level tasks across 11 domains like chip design and finance. Results show even top models like Claude Opus 5.5 average only 12%, with all models scoring zero in finance and law under strict grading, highlighting significant gaps in handling complex real-world workflows.

LATEST/SemiAnalysis released the GDPx benchmark to evaluate AI agents on expert-level tasks across 11 domains like chip design and finance. Results show even top models like Claude Opus 5.5 average only 12%, with all models scoring zero in finance and law under strict grading, highlighting significant gaps in handling complex real-world workflows.

Hourly heat curve

2 fully observed hours
Fewer than 3 fully observed hours; trend not plotted yet.

Coverage timeline

Total 1 reports · Latest first
  1. Rohan Paul (@rohanpaul_ai)T2·85 pts

    SemiAnalysis Launches GDPx Benchmark Showing Low AI Agent Scores on Expert Tasks

    • GDPx requires agents to autonomously navigate real workspaces with thousands of files, distinct from traditional context-injection evaluations.
    • Claude Opus 5.5 leads with a 12% average score, but all major models fail to pass strict grading standards in finance and law tasks.
    • In trading strategy improvement tasks, 70% of attempts performed worse than the original strategy, indicating models tend to increase risk rather than optimize logic.