ConflictQA - Text Conflict (Non-Complementary): leaderboard
Metric: Exact match (%) on 802 single-hop (non-complementary) questions when the provided evidence contains conflicting text passages (TextConf); zero-shot with a conflict-aware prompt, the mean of two evidence orders; exact match of the generated answer set; higher is better. Source: arxiv.org. Saturation forecast: Around 2032. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V3.2 (Non-reasoning) | 53.43 |
| 2 | O3 Mini | 52.56 |
| 3 | Qwen 3 8B | 52.44 |
| 4 | GPT-5.1 | 51.88 |
| 5 | Llama 3.1 70B Instruct | 48.96 |
| 6 | GPT-4o | 46.57 |
| 7 | GPT-3.5 Turbo (0125) | 43.77 |
| 8 | Llama 3.1 8B Instruct | 41.15 |
| 9 | DeepSeek V3.2 (Thinking) | 40.15 |
| 10 | Mistral Large 3 | 32.17 |
Interactive version: theaggregate.ai/benchmark?slug=conflictqa-text-conflict-non-complementary · How It Works · Data refreshed daily, snapshot 2026-10-07.