Graphwalks parents >128k: leaderboard
Long-context Graphwalks parent-retrieval task over 128k tokens, testing attention to graph structure across very large serialized inputs.
Metric: Score (%). Source: llm-stats.com. Status: saturated. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 | 95.4 |
| 2 | Claude Opus 4.8 | 83.3 |
| 3 | GPT-5.5 | 58.5 |
| 4 | GPT-5.4 | 32.4 |
| 5 | GPT-4.1 | 25 |
| 6 | GPT-4.1 Mini | 11 |
| 7 | GPT-4.1 Nano | 5.6 |
Interactive version: theaggregate.ai/benchmark?slug=graphwalks-parents-over-128k · How It Works · Data refreshed daily, snapshot 2026-09-05.