SAKE - Zero-Shot - Scalable Practices: leaderboard
Metric: Accuracy (%). Source: arxiv.org. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 93 |
| 2 | Claude Opus 4.6 | 92.7 |
| 3 | Grok 4.20 | 92.7 |
| 4 | Gemini 3.1 Flash Lite | 91.9 |
| 5 | GPT-5.4 | 91.1 |
| 6 | DeepSeek V3.2 | 91.1 |
| 7 | Qwen 3 235B A22B | 90.4 |
| 8 | Claude Haiku 4.5 | 89.6 |
| 9 | Mistral Small 4 | 89.6 |
| 10 | Grok 4.1 Fast | 88.4 |
| 11 | GPT-5.4 Nano | 88.4 |
Interactive version: theaggregate.ai/benchmark?slug=sake-zero-shot-scalable-practices · How It Works · Data refreshed daily, snapshot 2026-09-19.