HindiGen v1: leaderboard
Hindi generative-task benchmark for chat and instruct models, evaluated with the 3C3H rubric across Hindi QA, grammar, and safety tasks.
Metric: 3C3H Score (%). Source: huggingface.co. 35 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | O3 (2025-04-16) | 85.56 | #121 |
| 2 | O1 (2024-12-17) | 79.64 | #149 |
| 3 | Claude 3.5 Sonnet (20241022) | 77.47 | #291 |
| 4 | O4 Mini (2025-04-16) | 75.52 | #179 |
| 5 | Claude Opus 4 (20250514) | 74.49 | #154 |
| 6 | GPT-4o (2024-08-06) | 74.45 | #336 |
| 7 | GPT-4o (2024-05-13) | 73.98 | #325 |
| 8 | Gemini 2.5 Flash (Preview 05-20) | 73.65 | #216 |
| 9 | GPT-4.1 | 73.37 | #248 |
| 10 | GPT-4o (2024-11-20) | 72.44 | #381 |
| 11 | Gemini 2.5 Pro (Preview 05-06) | 71.77 | #147 |
| 12 | Claude 3.7 Sonnet (20250219) | 70.77 | #201 |
| 13 | Llama 3.1 70B Instruct | 70.45 | #561 |
| 14 | Claude Sonnet 4 (20250514) | 69.75 | #213 |
| 15 | GPT-4o Mini (2024-07-18) | 65.5 | #558 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=hindigen-v1 · How It Works · Data refreshed daily, snapshot 2026-10-20.