HELM SeaHELM - XQuAD (Vietnamese) — leaderboard
Metric: SQuAD macro-averaged F1 score. Source: crfm.stanford.edu. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.0 Flash | 69.86 |
| 2 | GPT-4o (2024-11-20) | 61.56 |
| 3 | Gemma 2 27B (IT) | 56.89 |
| 4 | Llama 3.2 90B Vision Instruct | 55.3 |
| 5 | Gemini 2.0 Flash Lite | 54.27 |
| 6 | Gemma 2 9B (IT) | 52.21 |
| 7 | Llama 4 Maverick Instruct FP8 | 50.87 |
| 8 | Claude 3.7 Sonnet (20250219) | 50.07 |
| 9 | Llama 4 Scout Instruct | 44.08 |
| 10 | Llama 3.3 70B Instruct | 43.43 |
| 11 | Llama 3.1 70B Instruct | 41.75 |
| 12 | GPT-4o Mini (2024-07-18) | 39.34 |
| 13 | DeepSeek V3 | 39.03 |
| 14 | Llama 3.1 405B Instruct | 38.93 |
| 15 | Qwen 2 72B Instruct | 38.92 |
Interactive version: theaggregate.ai/benchmark?slug=helm-seahelm-xquad-vietnamese · How the rankings work · Data refreshed daily, snapshot 2026-07-22.