Claude Opus 4.8 (Low): benchmark results
Provider: Anthropic. Released 2026-05-28. Access: API.
Aggregate Rating 1738 ± 23, rank #271 of 2171 rated models, from 30 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| FinLifeBench - Financial State - Checkpoint State Accuracy | 80.1 | Cell-level state accuracy (%) | 100 |
| FinLifeBench - Financial State - Granular Change Accuracy | 47 | GCA@15 (%) | 100 |
| K-Bench | 98.58 | Overall (%, Default v1 prompt) | 93.4 |
| OTIS Mock AIME 2024-25 | 97.78 | Accuracy (%) | 88.3 |
| K-Bench - Risk | 94.34 | Risk Score (%, Default v1 prompt) | 86.9 |
| Epoch AI - GPQA Diamond | 88.38 | Accuracy (%) | 79.5 |
| CellTrack-Bench - Inspection F1 | 0.5 | Score (0–1) | 76.2 |
| CellTrack-Bench - Mean F1 | 0.63 | Score (0–1) | 76.2 |
| CellTrack-Bench - Repair F1 | 0.73 | Score (0–1) | 76.2 |
| CellTrack-Bench - Subgraph LNK | 0.5 | Score (0–1) | 76.2 |
| K-Bench (Therapeutic v0) | 97.87 | Overall (%, Therapeutic v0 prompt) | 74.2 |
| ARC-AGI-2 | 62.22 | Accuracy (%) | 73.1 |
Interactive version: theaggregate.ai/model?slug=claude-opus-4-8-low · How It Works · Data refreshed daily, snapshot 2026-11-06.