HELM SeaHELM - IndicQA — leaderboard
Metric: SQuAD macro-averaged F1 score. Source: crfm.stanford.edu. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama 3.2 90B Vision Instruct | 57.23 |
| 2 | Llama 4 Maverick Instruct FP8 | 48.42 |
| 3 | Gemma 2 27B (IT) | 47.24 |
| 4 | Llama 3.1 70B Instruct | 43.31 |
| 5 | Llama 3.3 70B Instruct | 39.05 |
| 6 | GPT-4o (2024-11-20) | 35.44 |
| 7 | Llama 4 Scout Instruct | 34 |
| 8 | Llama 3.1 405B Instruct | 32.7 |
| 9 | Gemini 2.0 Flash Lite | 31.54 |
| 10 | Gemini 2.0 Flash | 29.76 |
| 11 | GPT-4o Mini (2024-07-18) | 29.39 |
| 12 | Qwen 2 72B Instruct | 27.21 |
| 13 | Gemma 2 9B (IT) | 26.25 |
| 14 | DeepSeek V3 | 22.69 |
| 15 | Qwen 2.5 72B Instruct | 19.16 |
Interactive version: theaggregate.ai/benchmark?slug=helm-seahelm-indicqa · How the rankings work · Data refreshed daily, snapshot 2026-07-22.