K-FinHallu - Four-Class Detection: leaderboard
Metric: Overall accuracy (%), micro-averaged over 808 test responses, at labelling each response faithful, hallucinated, false refusal or true refusal given the retrieved passages, multi-turn Korean financial RAG dialogues, temperature 0 where applicable; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 | 85.5 |
| 2 | Gemini 2.5 Pro | 85.3 |
| 3 | Gemini 2.5 Flash | 84.5 |
| 4 | Qwen 3 32B | 74.3 |
| 5 | Qwen 3 14B | 74.1 |
| 6 | Llama 3.1 70B | 69.3 |
| 7 | GPT-4o | 69.2 |
| 8 | Qwen 3 8B | 68.8 |
| 9 | Llama 3.3 70B | 68.7 |
| 10 | Llama 3.1 8B | 34.2 |
Interactive version: theaggregate.ai/benchmark?slug=k-finhallu-four-class-detection · How It Works · Data refreshed daily, snapshot 2026-10-07.