{"slug": "astra-opus-5-5-and-other-frontier-models-demonstrate-jagged-performance-across", "title": "Astra, Opus 5.5, and other Frontier Models Demonstrate Jagged Performance Across SoTA Agentic Tasks from Web Browsing to Robotics", "summary": "A team of researchers led by Yangyue Wang, with co-authors Harshvardhan Sikka, Pranav Guruprasad and Sudipta Chowdhury of Fig and the Georgia Institute of Technology, reported that frontier models including Astra and Opus 5.5 show jagged performance across state-of-the-art agentic tasks ranging from web browsing to robotics. The work evaluates these models on agentic benchmarks and finds uneven capability across task types.", "body_md": "Yangyue Wang\n1\n,\n    \nHarshvardhan Sikka\n1, 2\n,\n    \nPranav Guruprasad\n1\n,\n    \nSudipta Chowdhury\n1\n\n1\nFig\n;\n    \n2\nGeorgia Institute of Technology\n.", "url": "https://wpnews.pro/news/astra-opus-5-5-and-other-frontier-models-demonstrate-jagged-performance-across", "canonical_source": "https://www.fig.inc/blog/astra-opus-5-5-and-other-frontier-models-demonstrate-jagged-performance-across-sota-agentic-tasks-from-web-browsing-to-robotics/", "published_at": "2026-09-29 05:27:45+00:00", "updated_at": "2026-09-29 05:49:49.173241+00:00", "lang": "en", "topics": ["ai-agents", "ai-research", "large-language-models", "robotics", "artificial-intelligence"], "entities": ["Yangyue Wang", "Harshvardhan Sikka", "Pranav Guruprasad", "Sudipta Chowdhury", "Fig", "Georgia Institute of Technology", "Astra", "Opus 5.5"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/astra-opus-5-5-and-other-frontier-models-demonstrate-jagged-performance-across", "markdown": "https://wpnews.pro/news/astra-opus-5-5-and-other-frontier-models-demonstrate-jagged-performance-across.md", "text": "https://wpnews.pro/news/astra-opus-5-5-and-other-frontier-models-demonstrate-jagged-performance-across.txt", "jsonld": "https://wpnews.pro/news/astra-opus-5-5-and-other-frontier-models-demonstrate-jagged-performance-across.jsonld"}}