HELM WMT 2014: leaderboard
Machine translation quality on WMT 2014 benchmark across multiple language pairs, scored by BLEU-4. Part of Stanford HELM.
Metric: BLEU-4 (%). Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 91 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude 3 Opus (20240229) | 23.99 | #412 |
| 2 | Llama 3.1 405B Instruct | 23.81 | #457 |
| 3 | Gemini 1.5 Pro (002) | 23.13 | #349 |
| 4 | GPT-4o (2024-05-13) | 23.07 | #325 |
| 5 | Claude 3.5 Sonnet (20240620) | 22.9 | #299 |
| 6 | Nova Pro | 22.89 | #569 |
| 7 | Claude 3.5 Sonnet (20241022) | 22.56 | #291 |
| 8 | GPT-4o (2024-08-06) | 22.51 | #336 |
| 9 | Gemini 1.5 Flash (001) | 22.51 | #577 |
| 10 | Llama 3 70B | 22.46 | #631 |
| 11 | Llama 3.2 90B Vision Instruct | 22.41 | #622 |
| 12 | Llama 3.1 70B Instruct | 22.28 | #561 |
| 13 | Llama 3.3 70B Instruct | 21.92 | #534 |
| 14 | Claude 2 | 21.9 | #761 |
| 15 | Claude v1.3 | 21.85 | #1001 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-wmt-2014 · How It Works · Data refreshed daily, snapshot 2026-10-20.