HardcoreLogic - Uncommon Elements — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 | 67.88 |
| 2 | Grok 4 | 59.62 |
| 3 | GPT-5 Mini | 52.13 |
| 4 | GPT-OSS-120B | 51.11 |
| 5 | O4 Mini | 47.13 |
| 6 | Qwen 3 235B A22B (Thinking) | 40.94 |
| 7 | Grok 3 Mini | 39.5 |
| 8 | Gemini 2.5 Pro | 39.38 |
| 9 | DeepSeek V3.1 | 39.09 |
| 10 | Seed-OSS-36B-Instruct | 38.79 |
| 11 | DeepSeek R1 0528 | 37.28 |
| 12 | Qwen 3 Next 80B A3B (Thinking) | 32.13 |
| 13 | Claude Sonnet 4 (Thinking) | 28.25 |
| 14 | GLM-4.5 | 21.49 |
| 15 | Qwen 3 32B | 16.93 |
Interactive version: aibenchmarks.dev/benchmark?slug=hardcorelogic-uncommon-elements · How the rankings work · Data refreshed daily, snapshot 2026-07-17.