Open LLM Leaderboard v1 - ARC Challenge: leaderboard
Original Hugging Face Open LLM Leaderboard science questions, using ARC Challenge with 25 demonstrations and answer-length-normalized multiple-choice accuracy. Percentage, higher is better. Frozen v1 evaluation archive; distinct from the v2 tasks. Public test questions and possible training overlap limit comparisons with newer models.
Metric: Normalized accuracy (%) (25-shot). Source: huggingface.co. Saturation forecast: Estimated already saturated. 7252 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | free-evo-qwen72B-v0.8-re | 79.86 | #442 |
| 2 | Rhea-72B-v0.5 | 79.78 | #555 |
| 3 | Le_Triomphant-ECE-TW3 | 78.5 | #564 |
| 4 | luxia-21.4B-alignment-v1.2 | 77.73 | #963 |
| 5 | Smaug-72B-v0.1 | 76.02 | #613 |
| 6 | Myrrh_solar_10.7b_3.0 | 75.43 | #1060 |
| 7 | Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B | 74.91 | #848 |
| 8 | LLaMAntino-3-ANITA-8B-Inst-DPO-ITA | 74.57 | #1001 |
| 9 | UNA-SimpleSmaug-34B-v1beta | 74.57 | #716 |
| 10 | Luminex-34B-v0.2 | 74.49 | #692 |
| 11 | LogoS-7Bx2-MoE-13B-v0.2 | 74.4 | #873 |
| 12 | MoE_13B_DPO | 74.32 | #857 |
| 13 | DARE_TIES_13B | 74.32 | #856 |
| 14 | Smaug-34B-v0.1 | 74.23 | #721 |
| 15 | MixtureofMerges-MoE-4x7B-v5 | 73.89 | #899 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #6 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=open-llm-leaderboard-v1-arc-challenge · How It Works · Data refreshed daily, snapshot 2026-11-07.