{"slug": "gdpx-the-general-agentic-benchmark", "title": "GDPx: The General Agentic Benchmark", "summary": "A benchmark called GDPx v1.0 was published to measure frontier model capabilities across 11 domains, built with 14 partners and 206 tasks. The GDPx results are presented on a cost-versus-performance Pareto frontier plotting GDPx score against average cost per task on a log scale, with Muse Wonder Full excluded because it has no public price. The benchmark also lets users compare models by domain.", "body_md": "v1.0\n\nMeasuring frontier model capabilities across the most valuable domains\n\n- 11\n- domains\n- 14\n- partners\n- 206\n- tasks\n\n## Cost vs. Performance Pareto Frontier\n\n↑ GDPx scoreAvg cost per task (log) →\n\nNot shown: Muse Wonder Full, with no public price.\n\nby\n\n## Compare models by domain\n\nModelGDPx\n\nby\n\nThanks to", "url": "https://wpnews.pro/news/gdpx-the-general-agentic-benchmark", "canonical_source": "https://gdpx.semianalysis.com/", "published_at": "2026-11-02 05:55:35+00:00", "updated_at": "2026-11-02 06:21:42.744342+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-research"], "entities": ["GDPx", "Muse Wonder Full"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/gdpx-the-general-agentic-benchmark", "markdown": "https://wpnews.pro/news/gdpx-the-general-agentic-benchmark.md", "text": "https://wpnews.pro/news/gdpx-the-general-agentic-benchmark.txt", "jsonld": "https://wpnews.pro/news/gdpx-the-general-agentic-benchmark.jsonld"}}