LLM Stats (SQuALITY): leaderboard
Abstractive summarization of long narratives (short stories ~5K words) with multiple reference summaries per story testing comprehension and synthesis.
Metric: Score (%). Source: llm-stats.com. Saturation forecast: Estimated already saturated. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Phi-3.5-mini-instruct | 24.3 | #1167 |
| 2 | Phi-3.5-MoE-instruct | 24.1 | #700 |
| 3 | Nova Pro | 19.8 | #569 |
| 4 | Nova Lite | 19.2 | #754 |
| 5 | Nova Micro | 18.8 | #898 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-squality · How It Works · Data refreshed daily, snapshot 2026-10-20.