Benchmarks · agentic
GDPx
Uses partial credit, averaged equally across domains after averaging partners within each domain; excludes biology, cybersecurity, and alignment. Score: domain-balanced partial-credit score without safety classifiers. Collected by Jake Boggs' benchmark dashboard.
20 models · 20 results · data as of 2026-11-06 · gdpx.semianalysis.com ↗
Leaderboard
Opus 5.5 max
All results
20 rows| # | Model | Effort | GDPx |
|---|---|---|---|
| 1 | Opus 5.5 | max | 24.9% |
| 2 | GPT-6 Astra | max | 24.3% |
| 3 | GPT-6.1 Sol | max | 23.4% |
| 4 | Sonnet 5.5 | max | 20.7% |
| 5 | GPT-6.1 Terra | max | 20.6% |
| 6 | Gemini 4 Carbon | max | 19.3% |
| 7 | Fable 5.1 | max | 18.2% |
| 8 | Muse Wonder Full | max | 16.8% |
| 9 | Gemini 4 Argon | max | 12.7% |
| 10 | Muse Wonder (Max) | max | 12.3% |
| 11 | Haiku 5.5 | max | 11.6% |
| 12 | Grok 4.7 | max | 10.3% |
| 13 | Qwen 3.8 Max | max | 9.9% |
| 14 | DeepSeek V4.1 Flash 1017 | max | 9.7% |
| 15 | GLM 5.3 | max | 8.9% |
| 16 | GPT-6 Luna | max | 8.8% |
| 17 | Kimi K3 | max | 8.7% |
| 18 | MiniMax M3.1 | max | 8.4% |
| 19 | Inkling | max | 5.4% |
| 20 | Muse Spark Open | xhigh | 2.8% |