Machine Translation on WMT En-Zh 2023 (test)
80.7KIWI ScoreQwen-3 Chat (Q3-C-8B)
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Qwen-3 Chat (Q3-C-8B)Category=Chat, Scale=8B2026.01 | 80.7 | 63.53 | 88.35 | 86.3 | 82.31 | 85.77 | 14.23 | 98.18 | |
| Qwen-2.5 Chat (Q2.5-C-72B)Category=Chat, Scale=72B2026.01 | 80.57 | 65.13 | 89.22 | 86.78 | 82.76 | 85.99 | 14.6 | 98.11 | |
| Qwen2.5-72BCategory=Pre-trained, Scale=72B2026.01 | 80.4 | 66.25 | 89.8 | 86.94 | 82.82 | 86.15 | 14.85 | 98.32 | |
| Llama-3 Chat (L3-C-70B)Category=Chat, Scale=70B2026.01 | 80.02 | 63.42 | 88.36 | 86.07 | 81.88 | 84.79 | 14.6 | 98.03 | |
| Qwen2.5-7BCategory=Pre-trained, Scale=7B2026.01 | 79.47 | 64.03 | 88.73 | 85.97 | 82.27 | 85.18 | 15.16 | 98.03 | |
| Qwen-2.5 Chat (Q2.5-C-7B)Category=Chat, Scale=7B2026.01 | 79.27 | 62.08 | 87.75 | 85.4 | 81.92 | 85.11 | 15.2 | 98.05 | |
| Qwen-3 Reasoning (Q3-R-8B)Category=Reasoning, Scale=8B2026.01 | 79.21 | 62.38 | 87.25 | 84.62 | 81.37 | 84.26 | 15.43 | 96.77 | |
| MiniCPMCategory=Reasoning2026.01 | 78.71 | 62.86 | 88.29 | 84.94 | 82.07 | 84.74 | 13.37 | 97.92 | |
| Llama3-70BCategory=Pre-trained, Scale=70B2026.01 | 78.63 | 63.72 | 88.74 | 85.4 | 82.94 | 85.72 | 15.08 | 97.89 | |
| Llama-3 Chat (L3-C-8B)Category=Chat, Scale=8B2026.01 | 78.08 | 59.95 | 86.68 | 84.01 | 80.69 | 83.29 | 15.34 | 97.93 | |
| Llama3-8BCategory=Pre-trained, Scale=8B2026.01 | 77.12 | 61.15 | 87.46 | 83.73 | 81.88 | 84.24 | 14.81 | 97.32 | |
| Llama2-70BCategory=Pre-trained, Scale=70B2026.01 | 76.61 | 60.99 | 87.87 | 83.18 | 82.57 | 85.36 | 14.27 | 97.76 | |
| DeepSeek-671B (DS-671B)Category=Reasoning, Scale=671B2026.01 | 76.54 | 54.51 | 79.85 | 81.05 | 75.02 | 77.37 | 13.66 | 92.34 | |
| DeepSeek-Llama (DS-Llama)Category=Reasoning2026.01 | 75.79 | 57.68 | 85.53 | 81.64 | 80.77 | 82.21 | 13.42 | 96.21 | |
| mBART-50Category=NMT2026.01 | 75.24 | 58.97 | 86.32 | 80.81 | 82.44 | 84.09 | 13 | 97.8 | |
| DeepSeek-Qwen (DS-Qwen)Category=Reasoning2026.01 | 74.05 | 55 | 84.91 | 80.43 | 81.36 | 83.1 | 16.74 | 97.55 | |
| Llama2-7BCategory=Pre-trained, Scale=7B2026.01 | 71.96 | 54.39 | 84.63 | 78.77 | 79.31 | 79.89 | 14.02 | 93.96 | |
| NLLB-600MCategory=NMT2026.01 | 67.08 | 52.74 | 83.58 | 75.36 | 78.41 | 77.28 | 10.75 | 97.08 | |
| NLLB-3.3BCategory=NMT2026.01 | 66.06 | 52.11 | 83.13 | 75.72 | 75.82 | 73.78 | 10.62 | 96.07 | |
| NLLB-54BCategory=NMT2026.01 | 63.48 | 49.76 | 81.85 | 74.75 | 72.16 | 69.06 | 9.79 | 92.69 | |
| Llama-2 Chat (L2-C-7B)Category=Chat, Scale=7B2026.01 | 58.58 | 35.67 | 76.58 | 64.75 | 78.35 | 77.17 | 32.93 | 76.42 | |
| Llama-2 Chat (L2-C-70B)Category=Chat, Scale=70B2026.01 | 57.36 | 32.23 | 70.85 | 53.04 | 71.95 | 76.19 | 41.27 | 69.03 |