TruthfulQA: leaderboard
Truthfulness benchmark with 817 questions across 38 categories, catching models that imitate common human misconceptions or popular falsehoods.
Metric: Score (%). Source: arxiv.org. Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o Mini (2024-07-18) | 78.2 | #558 |
| 2 | Phi-3.5-MoE-instruct | 77.5 | #700 |
| 3 | Gemma 2 9B (IT) | 76.6 | #788 |
| 4 | Gemini 1.5 Flash | 76.6 | #618 |
| 5 | Llama 3.1 8B Instruct | 69.2 | #1035 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=truthfulqa · How It Works · Data refreshed daily, snapshot 2026-10-20.