Luth-1.7B-Instruct: benchmark results
Provider: Kurakura AI. Released 2025-06-01. Access: Open.
Aggregate Rating 1425 ± 23, rank #4916 of 10916 rated models, from 164 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Kurakurai French LLM - BFCL v2 | 71.05 | Accuracy (%) | 86.2 |
| Kurakurai French LLM - AIME 2024 | 10.33 | Accuracy (%) | 85.4 |
| Kurakurai French LLM - GPQA Diamond | 30.2 | Accuracy (%) | 85.4 |
| Kurakurai French LLM - MATH-500 | 65.26 | Accuracy (%) | 82.9 |
| Kurakurai French LLM - MMLU-Pro | 41.1 | Accuracy (%) | 82.9 |
| Kurakurai French LLM - AIME 2025 | 7 | Accuracy (%) | 80.5 |
| Kurakurai French LLM - HumanEval+ | 56.44 | pass@1 (%) | 80.5 |
| Kurakurai French LLM - Global MMLU | 58 | Accuracy (%) | 79.3 |
| EuroEval Bosnian NLU - Multi Wiki QA BS | 67.41 | Character F1 (%) | 78.8 |
| Kurakurai French LLM - MBPP | 56.4 | pass@1 (%) | 78 |
| Kurakurai French LLM - MGSM | 75.72 | Accuracy (%) | 78 |
| EuroEval Ukrainian NLU - MultiWikiQA UK | 59.67 | Character F1 (%) | 76.5 |
Interactive version: theaggregate.ai/model?slug=luth-1-7b-instruct · How It Works · Data refreshed daily, snapshot 2026-11-08.