Name: Claude Sonnet 5 vs Claude Sonnet 5.5
Creator: Benchmark Registry

Shared benchmarks

Shared benchmarks for Claude Sonnet 5 and Claude Sonnet 5.5
Benchmark	Claude Sonnet 5	Claude Sonnet 5.5
CursorBench 4.0	34.1%	53.1%4.0 · CursorBench Accuracy Cursor 35.8%4.0 · CursorBench Accuracy Cursor 55.5%4.0 · CursorBench Accuracy Cursor 39.2%4.0 · CursorBench Accuracy Cursor 47.8%4.0 · CursorBench Accuracy Cursor
Evaluation details for CursorBench 4.0 The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry. Claude Sonnet 5 Score 34.1% Evaluator Cursor Version CursorBench 4.0 Metric CursorBench Accuracy (percent) Reasoning max Reported September 10, 2026 Source Evaluation source for Claude Sonnet 5 on CursorBench 4.0 (opens in a new tab) Claude Sonnet 5.5 Score 53.1% Evaluator Cursor Version CursorBench 4.0 Metric CursorBench Accuracy (percent) Reasoning xhigh Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on CursorBench 4.0 (opens in a new tab) Score 35.8% Evaluator Cursor Version CursorBench 4.0 Metric CursorBench Accuracy (percent) Reasoning low Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on CursorBench 4.0 (opens in a new tab) Score 55.5% Evaluator Cursor Version CursorBench 4.0 Metric CursorBench Accuracy (percent) Reasoning max Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on CursorBench 4.0 (opens in a new tab) Score 39.2% Evaluator Cursor Version CursorBench 4.0 Metric CursorBench Accuracy (percent) Reasoning medium Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on CursorBench 4.0 (opens in a new tab) Score 47.8% Evaluator Cursor Version CursorBench 4.0 Metric CursorBench Accuracy (percent) Reasoning high Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on CursorBench 4.0 (opens in a new tab)
DeepSWE 1.1Potentially non-equivalent: Evaluator sets differ.	54.0%1.1 · DeepSWE Pass@1 DataCurve	71.0%1.1 · DeepSWE Pass@1 Anthropic
Potentially non-equivalent · Evaluation details for DeepSWE 1.1 Evaluator sets differ. Evaluation methodology is not recorded by the registry. Claude Sonnet 5 Score 54.0% Evaluator DataCurve Version DeepSWE 1.1 Metric DeepSWE Pass@1 (percent) Reasoning max Reported September 22, 2026 Source Evaluation source for Claude Sonnet 5 on DeepSWE 1.1 (opens in a new tab) Claude Sonnet 5.5 Score 71.0% Evaluator Anthropic Version DeepSWE 1.1 Metric DeepSWE Pass@1 (percent) Reasoning max Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on DeepSWE 1.1 (opens in a new tab)
Global MMLU Original	89.0%	92.1%
Evaluation details for Global MMLU Original The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry. Claude Sonnet 5 Score 89.0% Evaluator Anthropic Version Global MMLU Original Metric Average accuracy (percent) Reasoning max Reported June 30, 2026 Source Evaluation source for Claude Sonnet 5 on Global MMLU Original (opens in a new tab) Claude Sonnet 5.5 Score 92.1% Evaluator Anthropic Version Global MMLU Original Metric Average accuracy (percent) Reasoning max Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on Global MMLU Original (opens in a new tab)
Humanity's Last Exam Full set — No tools	43.2%	43.2%Full set — No tools · Humanity's Last Exam Accuracy Anthropic 56.9%Full set — No tools · Humanity's Last Exam Accuracy Anthropic 41.4%Full set — No tools · Humanity's Last Exam Accuracy Anthropic 47.9%Full set — No tools · Humanity's Last Exam Accuracy Anthropic 53.0%Full set — No tools · Humanity's Last Exam Accuracy Anthropic
Evaluation details for Humanity's Last Exam Full set — No tools The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry. Claude Sonnet 5 Score 43.2% Evaluator Anthropic Version Humanity's Last Exam Full set — No tools Metric Humanity's Last Exam Accuracy (percent) Reasoning max Reported June 30, 2026 Source Evaluation source for Claude Sonnet 5 on Humanity's Last Exam Full set — No tools (opens in a new tab) Claude Sonnet 5.5 Score 43.2% Evaluator Anthropic Version Humanity's Last Exam Full set — No tools Metric Humanity's Last Exam Accuracy (percent) Reasoning medium Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on Humanity's Last Exam Full set — No tools (opens in a new tab) Score 56.9% Evaluator Anthropic Version Humanity's Last Exam Full set — No tools Metric Humanity's Last Exam Accuracy (percent) Reasoning max Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on Humanity's Last Exam Full set — No tools (opens in a new tab) Score 41.4% Evaluator Anthropic Version Humanity's Last Exam Full set — No tools Metric Humanity's Last Exam Accuracy (percent) Reasoning low Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on Humanity's Last Exam Full set — No tools (opens in a new tab) Score 47.9% Evaluator Anthropic Version Humanity's Last Exam Full set — No tools Metric Humanity's Last Exam Accuracy (percent) Reasoning high Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on Humanity's Last Exam Full set — No tools (opens in a new tab) Score 53.0% Evaluator Anthropic Version Humanity's Last Exam Full set — No tools Metric Humanity's Last Exam Accuracy (percent) Reasoning xhigh Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on Humanity's Last Exam Full set — No tools (opens in a new tab)
Humanity's Last Exam Full set — With tools	47.2%Full set — With tools · Humanity's Last Exam Accuracy Anthropic 36.5%Full set — With tools · Humanity's Last Exam Accuracy Anthropic 54.6%Full set — With tools · Humanity's Last Exam Accuracy Anthropic 52.8%Full set — With tools · Humanity's Last Exam Accuracy Anthropic	64.5%Full set — With tools · Humanity's Last Exam Accuracy Anthropic 56.7%Full set — With tools · Humanity's Last Exam Accuracy Anthropic 50.5%Full set — With tools · Humanity's Last Exam Accuracy Anthropic 62.0%Full set — With tools · Humanity's Last Exam Accuracy Anthropic 47.6%Full set — With tools · Humanity's Last Exam Accuracy Anthropic
Evaluation details for Humanity's Last Exam Full set — With tools The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry. Claude Sonnet 5 Score 47.2% Evaluator Anthropic Version Humanity's Last Exam Full set — With tools Metric Humanity's Last Exam Accuracy (percent) Reasoning medium Reported June 30, 2026 Source Evaluation source for Claude Sonnet 5 on Humanity's Last Exam Full set — With tools (opens in a new tab) Score 36.5% Evaluator Anthropic Version Humanity's Last Exam Full set — With tools Metric Humanity's Last Exam Accuracy (percent) Reasoning low Reported June 30, 2026 Source Evaluation source for Claude Sonnet 5 on Humanity's Last Exam Full set — With tools (opens in a new tab) Score 54.6% Evaluator Anthropic Version Humanity's Last Exam Full set — With tools Metric Humanity's Last Exam Accuracy (percent) Reasoning xhigh Reported June 30, 2026 Source Evaluation source for Claude Sonnet 5 on Humanity's Last Exam Full set — With tools (opens in a new tab) Score 52.8% Evaluator Anthropic Version Humanity's Last Exam Full set — With tools Metric Humanity's Last Exam Accuracy (percent) Reasoning high Reported June 30, 2026 Source Evaluation source for Claude Sonnet 5 on Humanity's Last Exam Full set — With tools (opens in a new tab) Claude Sonnet 5.5 Score 64.5% Evaluator Anthropic Version Humanity's Last Exam Full set — With tools Metric Humanity's Last Exam Accuracy (percent) Reasoning max Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on Humanity's Last Exam Full set — With tools (opens in a new tab) Score 56.7% Evaluator Anthropic Version Humanity's Last Exam Full set — With tools Metric Humanity's Last Exam Accuracy (percent) Reasoning high Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on Humanity's Last Exam Full set — With tools (opens in a new tab) Score 50.5% Evaluator Anthropic Version Humanity's Last Exam Full set — With tools Metric Humanity's Last Exam Accuracy (percent) Reasoning medium Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on Humanity's Last Exam Full set — With tools (opens in a new tab) Score 62.0% Evaluator Anthropic Version Humanity's Last Exam Full set — With tools Metric Humanity's Last Exam Accuracy (percent) Reasoning xhigh Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on Humanity's Last Exam Full set — With tools (opens in a new tab) Score 47.6% Evaluator Anthropic Version Humanity's Last Exam Full set — With tools Metric Humanity's Last Exam Accuracy (percent) Reasoning low Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on Humanity's Last Exam Full set — With tools (opens in a new tab)
MILU Original	89.3%	91.6%
Evaluation details for MILU Original The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry. Claude Sonnet 5 Score 89.3% Evaluator Anthropic Version MILU Original Metric Average accuracy (percent) Reasoning max Reported June 30, 2026 Source Evaluation source for Claude Sonnet 5 on MILU Original (opens in a new tab) Claude Sonnet 5.5 Score 91.6% Evaluator Anthropic Version MILU Original Metric Average accuracy (percent) Reasoning max Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on MILU Original (opens in a new tab)
SWE-bench Multilingual	78.3%	90.3%
Evaluation details for SWE-bench Multilingual The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry. Claude Sonnet 5 Score 78.3% Evaluator Anthropic Version SWE-bench Multilingual Metric Resolved Tasks (percent) Reasoning max Reported June 30, 2026 Source Evaluation source for Claude Sonnet 5 on SWE-bench Multilingual (opens in a new tab) Claude Sonnet 5.5 Score 90.3% Evaluator Anthropic Version SWE-bench Multilingual Metric Resolved Tasks (percent) Reasoning max Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on SWE-bench Multilingual (opens in a new tab)
SWE-bench Pro Public	63.2%	81.3%
Evaluation details for SWE-bench Pro Public The recorded benchmark version, metric, and evaluator sets match. Evaluation methodology is not recorded by the registry. Claude Sonnet 5 Score 63.2% Evaluator Anthropic Version SWE-bench Pro Public Metric Public tasks resolved (percent) Reasoning max Reported June 30, 2026 Source Evaluation source for Claude Sonnet 5 on SWE-bench Pro Public (opens in a new tab) Claude Sonnet 5.5 Score 81.3% Evaluator Anthropic Version SWE-bench Pro Public Metric Public tasks resolved (percent) Reasoning max Reported September 28, 2026 Source Evaluation source for Claude Sonnet 5.5 on SWE-bench Pro Public (opens in a new tab)