HELM SeaHELM - IndicSentiment — leaderboard
Metric: Macro F1 score. Source: crfm.stanford.edu. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.0 Flash | 98.59 |
| 2 | Llama 3.1 405B Instruct | 98.3 |
| 3 | Gemini 2.0 Flash Lite | 98.24 |
| 4 | Llama 3.3 70B Instruct | 98.2 |
| 5 | Llama 4 Scout Instruct | 98.05 |
| 6 | Llama 3.2 90B Vision Instruct | 98 |
| 7 | GPT-4o Mini (2024-07-18) | 97.8 |
| 8 | Claude 3.7 Sonnet (20250219) | 97.75 |
| 9 | Llama 3.1 70B Instruct | 97.7 |
| 10 | DeepSeek V3 | 97.5 |
| 11 | GPT-4o (2024-11-20) | 97.3 |
| 12 | Qwen 2.5 72B Instruct | 93.89 |
| 13 | Claude 3.5 Haiku (20241022) | 93.48 |
| 14 | Qwen 2 72B Instruct | 75.29 |
| 15 | Gemma 2 27B (IT) | 45.34 |
Interactive version: theaggregate.ai/benchmark?slug=helm-seahelm-indicsentiment · How the rankings work · Data refreshed daily, snapshot 2026-07-22.