NLCO - Set-L - Accuracy: leaderboard
Metric: Accuracy: feasible and matching the solver-optimal objective (%). Source: arxiv.org. Saturation forecast: Around December 2026. 16 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Flash (High) | 60.9 |
| 2 | GPT-5.1 (Medium) | 57.2 |
| 3 | DeepSeek V3.2 (Thinking) | 54.1 |
| 4 | Grok 4.1 Fast (Reasoning) | 53.4 |
| 5 | O4 Mini (High) | 53.1 |
| 6 | Claude Sonnet 4.5 (Thinking) | 53 |
| 7 | Qwen 3 235B A22B 2507 Instruct | 36.3 |
| 8 | DeepSeek V3.2 (Non-reasoning) | 31.1 |
| 9 | Qwen 3 14B (Reasoning) | 25.5 |
| 10 | MiMo-V2-Flash (Non-reasoning) | 24.7 |
| 11 | QwQ-32B | 19.6 |
| 12 | Ministral-3-14B-Instruct-2512 | 12 |
| 13 | Llama 4 Maverick Instruct | 7.9 |
| 14 | Qwen 3 14B (Non-reasoning) | 7.8 |
Interactive version: theaggregate.ai/benchmark?slug=nlco-set-l-accuracy · How It Works · Data refreshed daily, snapshot 2026-09-25.