Evals for Every Language - Romanian (ro): leaderboard
Average score across evaluation tasks in Romanian (ro), from Evals for Every Language.
Metric: Average Score (%). Source: huggingface.co. Saturation forecast: Around January 2027. 67 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.7 | 78.5 | #49 |
| 2 | Gemini 3.1 Pro (Preview) | 77.99 | #58 |
| 3 | Claude Opus 4.8 | 77.92 | #37 |
| 4 | Gemini 3.1 Flash Lite | 77.78 | #191 |
| 5 | Claude Haiku 4.5 | 77.7 | #280 |
| 6 | Claude Sonnet 4.5 | 77.68 | #142 |
| 7 | DeepSeek V3.2 Exp | 76.29 | #234 |
| 8 | Grok 4.20 | 76.23 | #175 |
| 9 | GPT-5.1 | 76.2 | #135 |
| 10 | Claude Sonnet 4.6 | 75.81 | #90 |
| 11 | GPT-5.3 | 75.42 | #103 |
| 12 | Claude Sonnet 4 | 75.06 | #199 |
| 13 | GLM 4.5 Air | 75.05 | #392 |
| 14 | GPT-5 | 74.62 | #98 |
| 15 | Llama 4 Maverick | 74.47 | #464 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-ro · How It Works · Data refreshed daily, snapshot 2026-10-20.