HELM SeaHELM - XQuAD (Thai) — leaderboard
Metric: SQuAD macro-averaged F1 score. Source: crfm.stanford.edu. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama 4 Maverick Instruct FP8 | 77.14 |
| 2 | Llama 3.2 90B Vision Instruct | 72.81 |
| 3 | Llama 4 Scout Instruct | 71.11 |
| 4 | Llama 3.1 70B Instruct | 65.23 |
| 5 | Gemini 2.0 Flash Lite | 64.58 |
| 6 | Llama 3.3 70B Instruct | 64.44 |
| 7 | Gemini 2.0 Flash | 63.87 |
| 8 | Llama 3.1 405B Instruct | 62.41 |
| 9 | Gemma 2 27B (IT) | 59.03 |
| 10 | GPT-4o (2024-11-20) | 55.19 |
| 11 | Claude 3.7 Sonnet (20250219) | 49.78 |
| 12 | GPT-4o Mini (2024-07-18) | 43.41 |
| 13 | DeepSeek V3 | 42.46 |
| 14 | Qwen 2.5 72B Instruct | 38.64 |
| 15 | DeepSeek R1 Distill Llama 8B | 33.79 |
Interactive version: theaggregate.ai/benchmark?slug=helm-seahelm-xquad-thai · How the rankings work · Data refreshed daily, snapshot 2026-07-22.