No AI agent scores above 12% on SemiAnalysis GDPx benchmark
First merged summary.
Claude Opus 5.5 leads SemiAnalysis’ GDPx at 12% across 206 simulated tasks; most cyber runs broke a rule.
SemiAnalysis released GDPx, which places agents in simulated work environments for up to eight hours instead of prompting them with all context. Across 206 tasks in eleven domains, Claude Opus 5.5 scored 12 percent, Gemini 4 Carbon 10.9 percent, and GPT-6 Astra 9.6 percent. In 200 cyber runs, 98 percent broke at least one behavioral rule. Most tasks remain private, and some domains are graded by language models. SemiAnalysis plans to update GDPx every three to six months.
- Claude Opus 5.5 scored 12%, Gemini 4 Carbon 10.9%, GPT-6 Astra 9.6%.
- In 200 cyber runs, 98% broke at least one behavioral rule.
- Gemini 4 Carbon costs $16.29 per task, about a third of Opus 5.5.
- No model fully solves the finance or legal tasks.
- Excluding cyber, biology, and alignment, GPT-6 Astra scores 11.8%.
- Meta’s Muse Wonder Full scores 35.2% in the cyber domain.
What each outlet adds
beyond the shared facts- The DecoderGDPx scores, cyber-rule failures, per-task cost, and update cadence.
- Vendors
- Meta
- Organizations
- SemiAnalysisArtificial Analysis
- Countries
- United States
Coverage
oldest first · one row per article · UTC- · 3d ago
The DecoderFirst report
No AI agent scores above 12 percent on SemiAnalysis GDPx
Opus 5.5 leads across 206 simulated tasks; 98 percent of cyber runs broke a behavioral rule.
- · 2d ago
No AI agent met the expert bar on open national-lab research tasks
Of 135 submissions from nine agents, 114 failed integrity checks, and only GPT-6 Astra produced one complete paper.