Rombos-Coder-V2.5-Qwen-14B: benchmark results
A Qwen-family base/instruct TIES merge released by rombodawg using the creator's continuous finetuning recipe. It remains distinct from the upstream Qwen releases. Provider: rombodawg. Access: Open.
Aggregate Rating 1485 ± 33, rank #2597 of 10916 rated models, from 56 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Japanese LLM - Wiki PAS SET F1 | 11.44 | Score (%) | 94.5 |
| Open LLM Leaderboard - BBH | 61.65 | Score | 85.3 |
| Open LLM Leaderboard - MATH Level 5 | 33.01 | Score | 83.8 |
| Open LLM Leaderboard - IFEval | 70.47 | Score | 83.7 |
| Open Japanese LLM - Wiki Coreference SET F1 | 5.65 | Score (%) | 73.9 |
| Open LLM Leaderboard - MMLU-Pro | 39.39 | Score | 72.1 |
| Open Japanese LLM - Wiki NER SET F1 | 7.08 | Score (%) | 62.7 |
| Open LLM Leaderboard - GPQA | 30.29 | Score | 59.3 |
| Open Japanese LLM - RC | 90.63 | Score (%) | 58.7 |
| Open Japanese LLM - Janli Exact Match | 79.03 | Score (%) | 53.8 |
| Open Japanese LLM - MR | 86.4 | Score (%) | 51.7 |
| Open Japanese LLM - FA | 25.27 | Score (%) | 49 |
Interactive version: theaggregate.ai/model?slug=rombos-coder-v2-5-qwen-14b · How It Works · Data refreshed daily, snapshot 2026-11-08.