LLMEval-Fair - Engineering: leaderboard
Metric: Discipline Score (0-10). Source: github.com. 59 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek R1 | 9.47 |
| 2 | DeepSeek V3 | 9.3 |
| 3 | Kimi K2 | 9.23 |
| 4 | Qwen 3 235B A22B | 9.23 |
| 5 | Gemini 2.5 Pro (Preview) | 9.2 |
| 6 | O1 (2024-12-17) | 8.9 |
| 7 | Claude Sonnet 4.5 (Thinking) | 8.9 |
| 8 | GPT-5 | 8.83 |
| 9 | Claude Sonnet 4.5 | 8.8 |
| 10 | GLM-4.6 | 8.8 |
| 11 | DeepSeek V3.2 | 8.73 |
| 12 | Gemini 2.5 Flash (Thinking) | 8.67 |
| 13 | Claude Sonnet 4 | 8.57 |
| 14 | Claude Sonnet 4 (Thinking) | 8.57 |
| 15 | Qwen 3 32B | 8.43 |
Interactive version: theaggregate.ai/benchmark?slug=llmeval-fair-engineering · How It Works · Data refreshed daily, snapshot 2026-09-19.