Inverse Turing Bench: leaderboard
As AI systems integrate into online spaces, differentiating them from humans in conversations is increasingly important. We present Inverse Turing Bench, a benchmark that evaluates LLMs and other models on their ability to differentiate humans and AI in multi-turn text. The benchmark provides a collection of paired dialogue transcripts, wherein one dialogue is between two humans and the other is b
Metric: Accuracy (imported). Source: benchmarklist.com. Saturation forecast: Around September 2027. 29 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 (Adaptive Reasoning, Max Effort) | 77.92 | #63 (Claude Opus 4.6) |
| 2 | Claude Opus 4.6 (Non-reasoning) | 77.56 | #63 (Claude Opus 4.6) |
| 3 | GPT-5.5 (High) | 75.94 | #31 (GPT-5.5) |
| 4 | GPT-5.5 (Low) | 74.87 | #31 (GPT-5.5) |
| 5 | GPT-5.4 (High) | 68.22 | #79 (GPT-5.4) |
| 6 | GPT-5.1 (High) | 67.68 | #135 (GPT-5.1) |
| 7 | GPT-5.4 (Low) | 64.45 | #79 (GPT-5.4) |
| 8 | Claude Sonnet 4.6 (Non-reasoning) | 63.2 | #90 (Claude Sonnet 4.6) |
| 9 | GPT-5.1 (Low) | 63.02 | #135 (GPT-5.1) |
| 10 | Claude Opus 4.1 (Thinking) | 60.14 | #140 (Claude Opus 4.1) |
| 11 | Kimi K2.6 | 57.45 | #104 |
| 12 | Kimi K2.6 (Non-reasoning) | 55.3 | #104 (Kimi K2.6) |
| 13 | GPT-4 (0613) | 48.29 | #491 |
| 14 | Claude Haiku 4.5 (Non-reasoning) | 47.04 | #280 (Claude Haiku 4.5) |
| 15 | Qwen 3.6 35B A3B | 45.6 | #226 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=inverse-turing-bench · How It Works · Data refreshed daily, snapshot 2026-10-20.