Kaggle IndicGenBench - Flores-In: leaderboard
Flores-In from Google DeepMind's IndicGenBench on Kaggle: FLORES sentence translation between English and 29 Indic languages in both directions. Scored with character-level ChrF as in the IndicGenBench paper, 0-100, higher is better.
Metric: ChrF. Source: www.kaggle.com. Saturation forecast: Around September 2027. 12 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.5 | 59.58 | #148 |
| 2 | Claude Opus 4.1 | 56.71 | #146 |
| 3 | Gemini 3 Pro (Preview) | 56.66 | #71 |
| 4 | Gemini 2.5 Flash | 54.67 | #255 |
| 5 | GPT-5 | 51.15 | #105 |
| 6 | Qwen 3 235B A22B 2507 Instruct | 49.28 | #315 |
| 7 | O3 | 48.23 | #132 |
| 8 | GPT-4.1 | 42.47 | #258 |
| 9 | Grok 4 | 42.28 | #189 |
| 10 | O4 Mini | 41.88 | #185 |
| 11 | GPT-OSS-120B | 37.85 | #346 |
| 12 | Sarvam M | 36.85 | #643 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=kaggle-indicgenbench-flores-in · How It Works · Data refreshed daily, snapshot 2026-11-05.