Claude Sonnet 5 (Adaptive Reasoning, Max Effort): benchmark results

Claude Sonnet 5 evaluated in adaptive-reasoning mode at max effort. Provider: Anthropic. Released 2026-06-30. Access: API.

Unified ELO 1668 ± 1, rank #188 of 1761 rated models, from 24 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA Long Context Reasoning82Accuracy (%)94.9
UGI - Writing62.71Writing Score94.1
Artificial Analysis Intelligence Index45.11Intelligence Index93.8
AA Humanity's Last Exam41.29Accuracy (%)92.2
AA GPQA Diamond91.11Accuracy (%)91.7
UGI - Natural Intelligence53.88NatInt Score91.5
AA Omniscience16.45Score90.4
AA CritPt16.86Accuracy (%)90.1
AA Omniscience - Science, Engineering & Mathematics45.52Accuracy (%)88.3
AA-LCR77Accuracy (self-reported)87.4
AA GDPval1505.81ELO86.9
AA Omniscience - Software Engineering (SWE)62.7Accuracy (%)85.2

Interactive version: theaggregate.ai/model?slug=claude-sonnet-5-adaptive-reasoning-max-effort · How It Works · Data refreshed daily, snapshot 2026-09-05.