Claude Opus 5.5 (Medium): benchmark results
Provider: Anthropic. Released 2026-09-22. Access: API.
Aggregate Rating 1937 ± 13, rank #26 of 2201 rated models, from 85 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AI Coding Daily (Claude Code) - Python Code Quality | 18.5 | Python Code Quality (max 20) points, LLM-judged rubric score | 100 |
| DistDebug-Bench - Function | 60.4 | Coverage (%) | 100 |
| EmbodiedLab | 43.3 | T0 (raw control) success rate, 30 experiments (%) | 100 |
| EmbodiedLab - T0 Process | 64 | T0 (raw control) process score, six-milestone ladder (0-100) | 100 |
| EmbodiedLab - T1 Process | 73.7 | T1 (perception and motion planning) process score (0-100) | 100 |
| EmbodiedLab - T1 Success | 53.3 | T1 (perception and motion planning) success rate, 30 experim | 100 |
| MedSafe-Dx v0.3 (4aj) | 73.5 | Balanced escalation utility (fixed-flag policy = 0, perfect | 100 |
| QuantArena - One-shot - Overall | 42.61 | Score (%) | 100 |
| QuantArena - One-shot - Sealed Period | 40.65 | Score (%) | 100 |
| Supabase Evals (No Skills) | 89.9 | Tasks Passed (%) | 100 |
| Supabase Evals (Skills) | 92.8 | Tasks Passed (%) | 100 |
| nnInteractive Agents - Guided Radiology 9 Images | 0.83 | Mean Dice (0–1) | 100 |
Interactive version: theaggregate.ai/model?slug=claude-opus-5-5-medium · How It Works · Data refreshed daily, snapshot 2026-11-12.