{"cols": ["LFM3-1.7B", "LFM2.5-2.6B", "MiniCPM5-2B", "Gemma4-E4B", "Qwen3.5-4B", "Qwen3.5-9B"], "rows": [{"name": "LiveClaw", "v": [45.6, 14.7, 30.37, 20.36, 19.94, 35.77]}, {"name": "Claw-Eval General (EN), average", "v": [68.64, 62.63, 60.74, 56.14, 61.96, 66.53]}, {"name": "Claw-Eval General (EN), pass^3", "v": [51.4, 37.52, 34.58, 22.43, 42.99, 51.87]}, {"name": "BrowseComp-Plus (OpenClaw subset)", "v": [37.0, 26.33, 16.83, 21.0, 24.67, 38.0]}, {"name": "WebArena-Verified", "v": [38.42, 3.02, 9.11, 17.06, 21.92, 28.94]}, {"name": "EnterpriseOps-Gym", "v": [28.2, 7.94, 13.41, 9.86, 25.96, 28.97]}, {"name": "AutomationBench (public)", "v": [41.29, 0.94, 4.71, 1.92, 8.67, 15.1]}, {"name": "BFCL V4", "v": [71.39, 55.89, 62.65, 48.25, 55.12, 62.83]}, {"name": "ToolSandbox", "v": [80.36, 78.04, 68.16, 69.27, 75.4, 76.41]}, {"name": "MCP-Atlas", "v": [32.57, 16.45, 26.38, 6.19, 41.37, 42.67]}, {"name": "\u03c4\u00b3-bench (4 domains), average", "v": [68.01, 56.77, 69.44, 33.53, 65.98, 67.11]}, {"name": "SWE-bench Verified (OpenCode)", "v": [43.4, 18.0, 34.8, 0.5, 41.2, 51.2]}, {"name": "SWE-bench Multilingual (OpenCode)", "v": [23.0, 8.0, 15.67, 0.67, 23.33, 35.0]}, {"name": "Terminal-Bench 2.1 (OpenCode)", "v": [12.92, 0.0, 5.62, 0.0, 13.48, 17.98]}, {"name": "LiveCodeBench v6", "v": [73.46, 60.57, 70.52, 71.0, 61.99, 69.86]}, {"name": "LiveCodeBench Pro 25Q2", "v": [31.74, 19.16, 29.34, 29.94, 24.25, 34.43]}, {"name": "SciCode", "v": [28.47, 16.32, 27.78, 25.0, 23.02, 31.6]}, {"name": "IFEval", "v": [96.66, 95.39, 87.83, 90.02, 92.52, 92.08]}, {"name": "IFBench", "v": [73.03, 60.48, 62.08, 44.56, 38.16, 48.84]}, {"name": "IFStruct", "v": [93.72, 82.18, 72.65, 78.45, 87.05, 89.2]}, {"name": "Multi-IF", "v": [89.1, 84.8, 73.19, 81.03, 80.32, 84.44]}, {"name": "GPQA Diamond", "v": [63.37, 57.78, 65.15, 57.88, 75.76, 80.4]}, {"name": "Humanity's Last Exam (no tools)", "v": [7.17, 5.88, 8.2, 4.77, 12.84, 16.41]}, {"name": "Humanity's Last Exam (with tools)", "v": [13.9, 9.6, 11.5, 6.0, 11.6, 17.5]}, {"name": "AA-Omniscience (public)", "v": [0.5, -33.17, -5.0, -15.33, -57.17, -28.0]}, {"name": "AA-Omniscience accuracy (public)", "v": [7.07, 8.75, 9.13, 9.83, 15.67, 13.17]}, {"name": "AA-Omniscience non-hallucination (public)", "v": [92.93, 53.69, 84.44, 72.09, 13.64, 52.59]}, {"name": "AA-LCR", "v": [52.83, 6.0, 49.34, 39.33, 52.67, 62.0]}, {"name": "RULER 128K", "v": [65.56, 1.15, 37.96, 18.82, 84.63, 87.11]}, {"name": "Average", "v": [49.37, 31.55, 39.19, 31.74, 39.27, 47.24]}]}