HellaSwag
Rank #43 of 51 · 52 models scored · top score 93.73
unknown
Strongest on discrimination, weakest on contamination resistance. Caveat: scores come from mixed harnesses, so a head-to-head is not apples-to-apples.
Weighted composite. Each signal is computed from the source-backed leaderboard below.
Does it still separate models?
How far from ceiling / clustered at the top?
Public vs held-out; training-leak risk.
Apples-to-apples, or mixed / vendor-optimized?
How old is the benchmark?
| # | Model | Score | Evidence | Measured |
|---|---|---|---|---|
| 1 | GPT-4 (Mar 2023) | 93.73 | unverified· optimized | 2023-03-15 |
| 2 | Llama 3.1-405B | 85.6 | self-reported· optimized | 2024-07-23 |
| 3 | Falcon-180B | 85.33 | unverified· optimized | 2023-09-06 |
| 4 | DeepSeek-V3 | 85.2 | self-reported· optimized | 2024-12-24 |
| 5 | DeepSeek-V2 (MoE-236B, May 2024) | 82.8 | self-reported· optimized | 2024-05-07 |
| 6 | PaLM 2-L | 82.4 | self-reported· optimized | 2023-05-17 |
| 7 | Mixtral 8x7B | 82.27 | unverified· optimized | 2023-12-11 |
| 8 | Llama 2-70B | 80.4 | self-reported· optimized | 2023-07-18 |
| 9 | Falcon-40B | 80.37 | self-reported· optimized | 2023-03-15 |
| 10 | Qwen2.5-72B | 79.73 | self-reported· optimized | 2024-09-19 |
| 11 | LLaMA-65B | 78.93 | unverified· optimized | 2023-02-24 |
| 12 | Stable Beluga 2 | 78.8 | self-reported· optimized | 2023-07-20 |
| 13 | PaLM 2-M | 78.67 | self-reported· optimized | 2023-05-17 |
| 14 | Qwen2.5-Coder-32B | 77.33 | self-reported· optimized | 2024-09-18 |
| 15 | Falcon 2 11B | 77.21 | self-reported· optimized | 2024-05-09 |
| 16 | LLaMA-33B | 77.07 | unverified· optimized | 2023-02-27 |
| 17 | Nemotron-4 15B | 76.53 | self-reported· optimized | 2024-02-27 |
| 18 | phi-3-medium 14B | 76.53 | self-reported· optimized | 2024-04-23 |
| 19 | Gemma 7B | 76.27 | unverified· optimized | 2024-02-21 |
| 20 | PaLM 2-S | 76 | self-reported· optimized | 2023-05-17 |
| 21 | Mistral 7B v0.1 | 74.67 | unverified· optimized | 2023-10-10 |
| 22 | Llama 2-13B | 74.27 | unverified· optimized | 2023-07-18 |
| 23 | Qwen2.5-Coder-14B | 73.6 | self-reported· optimized | 2024-09-18 |
| 24 | LLaMA-13B | 72.27 | unverified· optimized | 2023-02-27 |
| 25 | internlm-20b | 70.8 | self-reported· optimized | 2023-09-18 |
| 26 | Falcon-7B | 70.8 | self-reported· optimized | 2023-04-24 |
| 27 | Llama 2-7B | 69.6 | unverified· optimized | 2023-07-18 |
| 28 | phi-3-small 7.4B | 69.33 | self-reported· optimized | 2024-04-23 |
| 29 | Qwen2.5-Coder (7B) | 69.07 | self-reported· optimized | 2024-09-18 |
| 30 | phi-3-mini 3.8B | 68.93 | self-reported· optimized | 2024-04-23 |
| 31 | MPT-7B | 68.53 | self-reported· optimized | 2023-05-05 |
| 32 | Yi-9B | 68.53 | unverified· optimized | 2024-03-01 |
| 33 | LLaMA-7B | 68.27 | unverified· optimized | 2023-02-24 |
| 34 | Yi 6B | 65.87 | unverified· optimized | 2023-11-02 |
| 35 | XGen-7B | 65.6 | self-reported· optimized | 2023-09-07 |
| 36 | open_llama_7b | 62.4 | self-reported· optimized | 2023-06-07 |
| 37 | INTELLECT-1 | 61.89 | self-reported· optimized | 2024-11-29 |
| 38 | Gemma 2B | 61.87 | unverified· optimized | 2024-02-21 |
| 39 | Qwen2.5-Coder-3B | 61.2 | self-reported· optimized | 2024-09-18 |
| 40 | Baichuan2-13B | 61.07 | self-reported· optimized | 2023-09-06 |
| 41 | Dolly 2.0-12b | 61.07 | self-reported· optimized | 2023-04-12 |
| 42 | internlm-7b | 60.8 | self-reported· optimized | 2023-07-05 |
| 43 | RedPajama-INCITE-7B-Base | 60.4 | self-reported· optimized | 2023-05-04 |
| 44 | Baichuan 2-7B | 57.33 | self-reported· optimized | 2023-09-20 |
| 45 | Qwen2.5-Coder (1.5B) | 49.07 | self-reported· optimized | 2024-09-18 |
| 46 | Cerebras-GPT-13B | 45.87 | self-reported· optimized | 2023-04-06 |
| 47 | vicuna-13b-v1.1 | 43.73 | self-reported· optimized | 2023-04-12 |
| 48 | chatglm2-6b | 42.67 | self-reported· optimized | 2023-06-24 |
| 49 | Phi-2 | 38.13 | self-reported· optimized | 2023-12-12 |
| 50 | Qwen2.5-Coder-0.5B | 31.2 | self-reported· optimized | 2024-09-18 |
| 51 | Phi-1.5 | 30.13 | self-reported· optimized | 2023-09-11 |
| 52 | stablelm-tuned-alpha-7b | 20.93 | self-reported· optimized | 2023-04-19 |