General Language Understanding and Reasoning on Open LLM Leaderboard Lighteval (test)
91.07Mean AccuracyGPT-5
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| GPT-5Model Category=Larger Instruct and Proprietary Model2026.01 | 91.07 | 91.4 | 95.31 | 91.36 | 94.85 | 87.1 | 87.85 | 89.6 | |
| RedSage-8B-DPOModel Category=Instruct Model, Alignment=DPO, Parameters=8B2026.01 | 74.33 | 77.07 | 71.76 | 82.71 | 79.87 | 52.47 | 73.01 | 83.44 | |
| RedSage-8B-InsModel Category=Instruct Model, Alignment=Instruction Tuned, Parameters=8B2026.01 | 73.34 | 77.38 | 69.62 | 86.05 | 79 | 47.75 | 73.64 | 79.97 | |
| Qwen3-32BModel Category=Larger Instruct and Proprietary Model, Parameters=32B2026.01 | 73.17 | 82.11 | 69.28 | 87.49 | 70.93 | 48.17 | 65.98 | 88.26 | |
| Qwen3-8B-BaseModel Category=Base Model, Parameters=8B2026.01 | 70.86 | 78.73 | 68.09 | 81.73 | 79.62 | 43.84 | 73.16 | — | |
| RedSage-8B-SeedModel Category=Base Model, Strategy=Seed, Parameters=8B2026.01 | 69.58 | 78.18 | 65.19 | 82.34 | 77.76 | 42.44 | 71.59 | — | |
| RedSage-8B-CFWModel Category=Base Model, Strategy=CFW, Parameters=8B2026.01 | 69.31 | 78.63 | 66.72 | 81.12 | 79.26 | 38.09 | 72.06 | — | |
| Foundation-Sec-8B-InstructModel Category=Instruct Model, Parameters=8B2026.01 | 69.28 | 64.11 | 63.91 | 77.79 | 81.35 | 53.15 | 68.51 | 76.17 | |
| RedSage-8B-BaseModel Category=Base Model, Parameters=8B2026.01 | 69.23 | 77.8 | 65.53 | 82.03 | 77.96 | 42.19 | 69.85 | — | |
| Llama-3.1-8B-InstructModel Category=Instruct Model, Parameters=8B2026.01 | 68.2 | 67.29 | 57.51 | 77.41 | 78.91 | 45.93 | 72.61 | 77.75 | |
| Llama-Primus-MergedModel Category=Instruct Model, Parameters=8B2026.01 | 66.71 | 66.17 | 53.07 | 75.28 | 79.07 | 46.52 | 73.24 | 73.58 | |
| Qwen3-8BModel Category=Instruct Model, Parameters=8B2026.01 | 65.92 | 73.59 | 62.54 | 75.66 | 56.7 | 45.23 | 62.51 | 85.21 | |
| DeepHat-V1-7BModel Category=Instruct Model, Parameters=7B2026.01 | 64.89 | 69.53 | 57.17 | 77.94 | 74.8 | 33.17 | 69.06 | 72.58 | |
| Llama-Primus-BaseModel Category=Instruct Model, Parameters=8B2026.01 | 64.82 | 65.09 | 51.19 | 71.8 | 79.49 | 44.62 | 72.69 | 68.85 | |
| Llama-3.1-8BModel Category=Base Model, Parameters=8B2026.01 | 61.15 | 66.31 | 58.19 | 49.05 | 82.08 | 35.98 | 75.3 | — | |
| Foundation-Sec-8BModel Category=Base Model, Parameters=8B2026.01 | 60.24 | 63.62 | 58.45 | 46.17 | 81.32 | 38.71 | 73.16 | — | |
| Lily-Cybersecurity-7B-v0.2Model Category=Instruct Model, Parameters=7B2026.01 | 56.98 | 56.49 | 58.96 | 30.86 | 80.94 | 48.53 | 72.06 | 50.99 |