Claude Opus 5.5 (High): benchmark results
Provider: Anthropic. Released 2026-09-22. Access: API.
Aggregate Rating 1941 ± 10, rank #22 of 2201 rated models, from 168 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AI Coding Daily (Claude Code) - Total | 85.43 | Total points (max 90) | 100 |
| GenData Experiment Selection - Headline 281 - Negative Records | 0.56 | Paired Accuracy (0–1) | 100 |
| GenData Experiment Selection - Headline 281 - Positive Records | 0.37 | Paired Accuracy (0–1) | 100 |
| MathArena - ARXIVLEAN September | 47.17 | Accuracy (%) | 100 |
| MathArena Arxiv False | 95 | Accuracy (%) | 100 |
| ParseBench - Tables | 94.26 | GTRM Score (GriTS + TableRecordMatch) | 100 |
| SelfBench - PostHog/posthog | 68.57 | Task success (%) on 35 tasks rebuilt from the repository's m | 100 |
| SelfBench - pingdotgg/t3code | 76.92 | Task success (%) on 39 tasks rebuilt from the repository's m | 100 |
| SelfBench - vercel/vercel | 80.85 | Task success (%) on 47 tasks rebuilt from the repository's m | 100 |
| Senior SWE-Bench | 39.33 | Tasteful Solve Rate (pass@1, %) | 100 |
| SweetBench - COGI | 73.54 | Control score (0–100) | 100 |
| SweetBench - Glucose variability subscore | 81.28 | Control score (0–100) | 100 |
Interactive version: theaggregate.ai/model?slug=claude-opus-5-5-high · How It Works · Data refreshed daily, snapshot 2026-11-12.