BhashaBench - Finance: leaderboard
From BharatGen's BhashaBench V1 leaderboard, which scores models on English and Hindi questions sourced from Indian government and domain-specific exams. BhashaBench-Finance tests Indian finance and economics knowledge; this board is accuracy over all questions in both languages (%), higher is better. Flagged runs are skipped.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Around April 2027. 31 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-OSS-120B | 71.05 | #346 |
| 2 | DeepSeek V3.1 | 61.48 | #275 |
| 3 | Qwen 3 235B A22B 2507 Instruct | 61.43 | #315 |
| 4 | GPT-4o | 54.97 | #357 |
| 5 | Gemma 2 27B | 45.77 | #803 |
| 6 | Gemma 2 9B | 40.94 | #924 |
| 7 | Gemma 2 27B (IT) | 39.95 | #652 |
| 8 | Pangea-7B | 39.25 | #1223 |
| 9 | Qwen 2.5 3B | 37.26 | #1177 |
| 10 | Gemma 2 9B (IT) | 36.75 | #795 |
| 11 | GPT-OSS-20B | 35.73 | #523 |
| 12 | Indic-gemma-7B-finetuned-sft-Navarasa-2.0 | 34.99 | #1372 |
| 13 | Llama 3.1 8B | 34.48 | #1130 |
| 14 | Llama 3.1 8B Instruct | 34.01 | #1017 |
| 15 | aya-23-8B | 33.9 | #1267 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=bhashabench-finance · How It Works · Data refreshed daily, snapshot 2026-11-05.