SemiAnalysis launches GDPx: best frontier model scores just 12% on expert work
rohanpaul_ai · x · 2026-11-02
SemiAnalysis released GDPx, an agentic benchmark spanning 11 high-value domains (chip design, cyber, finance, law, quant trading, RSI, and more) built with 14 data partners. Claude Opus 5.5 tops it at just 12%, Gemini 4 Carbon follows at 10.9%. Unlike GDPval, agents must dig through real workspaces (e.g., a 3,983-file acquisition replica). Excluding domains where safety classifiers intervene lifts GPT-6 Astra to 11.8%, past Opus. Finance and law defeat every model under pass-fail grading, and models worsened a working trading strategy in 70 of 111 attempts, mostly by taking bigger risks.
Related event: SemiAnalysis Launches GDPx Frontier Benchmark; Best Model Scores Just 12%(10 posts)→
More from Models
- Fable 5.5 creates 66-second animated short entirely with self-written code, no image models — oran_ge · 2026-11-02
- SenseTime open-sources SenseNova-Loop-9B: weight-shared 56-layer loop, 256K context, Apache 2.0 — AdinaYakup · 2026-11-02
- User says Gemini excels at long-form fanfic, producing stories up to 29 minutes of reading — Ok-Jello3739 · 2026-11-02
- Sam Altman calls tokens 'a kind of fake unit' that doesn't measure business value — rohanpaul_ai · 2026-11-02
- 'You Can Feel the Gradient Slopes': Dev Contrasts Benchmaxxed RL Models vs General Ones — theteknosaur · 2026-11-02
- LightOnOCR-3 Launches: Tops FRBench, Adds Layout Structure and Chart Extraction — IgorCarron · 2026-11-02