Literary Translation on MetaphorTrans (in-domain)
79.95CRF ScoreClaude Opus 4.5
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Claude Opus 4.52026.06 | 79.95 | 3.9 | 73.3 | |
| Kimi K2Params=1T (32B), thinking-enabled=true2026.06 | 79.93 | 3.81 | 71.7 | |
| Claude Sonnet 4.52026.06 | 77.46 | 3.66 | 68.43 | |
| GPT-5.22026.06 | 77.45 | 3.68 | 68.68 | |
| DeepSeek V3.1Params=671B (37B)2026.06 | 77.19 | 3.75 | 70.27 | |
| LitMT-14BParams=14B, trained on MetaphorTrans=true2026.06 | 75.2 | 3.71 | 69.07 | |
| Qwen3-235B-A22BParams=235B (22B)2026.06 | 74.15 | 3.55 | 65.62 | |
| LitMT-8BParams=8B, trained on MetaphorTrans=true2026.06 | 73.03 | 3.61 | 67.25 | |
| ExTrans-7BParams=7B, trained on MetaphorTrans=true, thinking-enabled=true2026.06 | 71.87 | 3.42 | 62.95 | |
| DeepTrans-7BParams=7B, trained on MetaphorTrans=true, thinking-enabled=true2026.06 | 70.31 | 3.33 | 61.15 | |
| DRT-14BParams=14B, trained on MetaphorTrans=true, thinking-enabled=true2026.06 | 67.07 | 3.21 | 58.43 | |
| Qwen3-32BParams=32B2026.06 | 66.65 | 3.15 | 57.09 | |
| Qwen3-14BParams=14B2026.06 | 65.25 | 3.07 | 55.47 | |
| MiniMax M2Params=230B (10B)2026.06 | 65.13 | 2.96 | 53.15 | |
| DRT-7BParams=7B, trained on MetaphorTrans=true, thinking-enabled=true2026.06 | 64.48 | 3.05 | 55.09 | |
| Qwen3-8BParams=8B2026.06 | 62.7 | 2.94 | 52.77 | |
| gpt-oss-120BParams=120B (5.1B), thinking-enabled=true2026.06 | 62.25 | 2.89 | 51.5 | |
| Llama3.3-70BParams=70B2026.06 | 59.46 | 2.75 | 48.34 | |
| gpt-oss-20BParams=20B (3.6B), thinking-enabled=true2026.06 | 55.97 | 2.59 | 44.7 | |
| LMT-60-8BParams=8B2026.06 | 54.59 | 2.59 | 44.63 | |
| X-ALMA-13B (G6)Params=13B2026.06 | 51.35 | 2.44 | 41.11 |