GaelEval - Gaelic to English Translation (chrF): leaderboard
GaelEval (2026) is a Scottish Gaelic benchmark from the DR-LIB project. This board is chrF, a character n-gram F-score (0 to 100), for translating 908 An Litir Bheag podcast transcripts from Gaelic into English, scored against the professional parallel translations; higher is better.
Metric: chrF character n-gram F-score of Gaelic-to-English translation against gold references; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 18 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 78.2 | #119 |
| 2 | DeepSeek R1 | 77.73 | #482 |
| 3 | Gemini 2.5 Flash | 77.58 | #463 |
| 4 | Claude Opus 4.6 | 77.38 | #113 |
| 5 | Gemini 3 Flash (Preview) | 77.38 | #152 |
| 6 | GPT-5.2 | 76.18 | #185 |
| 7 | GPT-5 | 75.82 | #178 |
| 8 | Llama 4 Maverick | 75.74 | #1110 |
| 9 | GPT-4.1 Mini | 75.48 | #789 |
| 10 | GPT-4o | 74.8 | #751 |
| 11 | Claude Haiku 4.5 | 73.92 | #583 |
| 12 | GPT-5 Mini | 73.34 | #364 |
| 13 | GPT-4o Mini | 73.29 | #1667 |
| 14 | GPT-4.1 | 72.77 | #505 |
| 15 | GPT-4.1 Nano | 69.04 | #2056 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=gaeleval-gaelic-to-english-translation-chrf · How It Works · Data refreshed daily, snapshot 2026-11-21.