HumanLikeness - Overall: leaderboard
Psycholinguistic humanlikeness benchmark comparing model response distributions with 2,000 human participants across sound, word, syntax, meaning, and discourse tasks.
Metric: Overall Humanlike (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 20 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 3.1 70B Instruct | 66.91 | #561 |
| 2 | Llama 3.1 8B Instruct | 66.2 | #1035 |
| 3 | Phi-3 Mini 4K Instruct | 64.73 | #1297 |
| 4 | Mistral Nemo Instruct (2407) | 63.97 | #936 |
| 5 | Llama 2 13B Chat Base | 63.26 | #1422 |
| 6 | CodeLlama-34B-Instruct-hf | 62.16 | #1381 |
| 7 | c4ai-command-r-plus | 61.09 | #724 |
| 8 | Llama 3 8B Instruct | 61.01 | #1129 |
| 9 | starchat2-15B-v0.1 | 60.84 | #832 |
| 10 | GPT-4o | 58.98 | #343 |
| 11 | GPT-3.5 Turbo | 58.66 | #865 |
| 12 | Yi 1.5 34B Chat | 58.5 | #800 |
| 13 | Llama 2 7B Chat | 57.6 | #1499 |
| 14 | Llama 3 70B Instruct | 57.14 | #634 |
| 15 | zephyr-7B-alpha | 56.89 | #1312 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=humanlikeness-overall · How It Works · Data refreshed daily, snapshot 2026-10-20.