SeaEval - Cross-Lingual Consistency - Cross-LogiQA (Zero-Shot): leaderboard
SeaEval zero-shot Cross-LogiQA evaluation measuring answer accuracy and cross-lingual consistency across English, Chinese, Spanish, Vietnamese, Indonesian, Malay, and Filipino.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 23 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 72B Instruct | 72.48 | #454 |
| 2 | Qwen 2.5 32B Instruct | 69.32 | #498 |
| 3 | Qwen 2 72B Instruct | 67.29 | #576 |
| 4 | Llama 3.1 70B Instruct | 65.67 | #561 |
| 5 | Qwen 2.5 14B Instruct | 64.37 | #647 |
| 6 | Llama 3 70B Instruct | 63.07 | #634 |
| 7 | Gemma 2 9B (IT) | 61.85 | #788 |
| 8 | gemma2-9B-cpt-sea-lionv3-instruct | 60.55 | #705 |
| 9 | Qwen 2.5 7B Instruct | 59.9 | #862 |
| 10 | Qwen 2 7B Instruct | 56.49 | #991 |
| 11 | SeaLLMs-v3-7B-Chat | 55.52 | #915 |
| 12 | Sailor2-8B-Chat | 54.06 | #782 |
| 13 | Llama 3 8B Cpt Sea Lionv2.1 Instruct | 49.92 | #1024 |
| 14 | Qwen 2.5 3B Instruct | 48.78 | #1162 |
| 15 | Llama 3.1 8B Instruct | 48.62 | #1035 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=seaeval-cross-lingual-consistency-cross-logiqa-zero-shot · How It Works · Data refreshed daily, snapshot 2026-10-20.