Claude Opus 4 (Thinking): benchmark results
Claude Opus 4 evaluated with thinking enabled. Provider: Anthropic. Released 2025-05-22. Access: API.
Unified ELO 1630 ± 1, rank #318 of 1761 rated models, from 31 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| BenchTable | 82.8 | Total Score (%) | 99.1 |
| Ducky Bench (Saxo Frog) | 1280 | ELO | 95.8 |
| Wolfram LLM Benchmarking Project | 62.4 | Correct Functionality (%) | 91.2 |
| Kagi LLM Benchmark | 74.3 | Accuracy (%) | 90.1 |
| SEAL - MASK | 87.87 | Score | 86.4 |
| ZEROBench-Sub | 25.06 | Score (self-reported) | 75 |
| AA Terminal-Bench Hard | 31.06 | Accuracy (%) | 73.7 |
| TrackingAI IQ Test (Offline) | 75 | Offline IQ Score (%) | 72.4 |
| Artificial Analysis Intelligence Index | 24.28 | Intelligence Index | 71.5 |
| Ducky Bench (Stabby Quack) | 1197 | ELO | 71.4 |
| REAL Evals | 19.7 | Task Completion (%) | 67.9 |
| SEAL - VISTA | 46.96 | Score | 67.7 |
Interactive version: theaggregate.ai/model?slug=claude-opus-4-thinking · How It Works · Data refreshed daily, snapshot 2026-09-05.