Qwen3.8-27B-NVFP4: benchmark results
Provider: Alibaba. Access: Open.
Aggregate Rating 1701 ± 25, rank #194 of 1635 rated models, from 28 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| llmbenchmarks.io Local GPU - IFBench | 69.77 | Instruction-level strict accuracy (%) | 100 |
| llmbenchmarks.io Local GPU - IFEval | 92.79 | Prompt-level strict accuracy (%) | 95.7 |
| BenCzechMark - Reading Comprehension | 76.5 | Duel Win Score (%) | 93.6 |
| llmbenchmarks.io Local GPU - ARC-AGI-1 Public Eval | 12.38 | pass@2 on the public evaluation set (%) | 90.9 |
| llmbenchmarks.io Local GPU - SciCode | 35.65 | Subproblems solved (%), pass@1 averaged over 3 epochs | 81.8 |
| BenCzechMark - Czech Language Understanding | 66.67 | Duel Win Score (%) | 79.5 |
| BenCzechMark - Czech Math Reasoning | 62.5 | Duel Win Score (%) | 75.6 |
| BenCzechMark - NER | 72.44 | Duel Win Score (%) | 73.1 |
| llmbenchmarks.io Local GPU - Chess Puzzles | 5 | Puzzles solved (%), exact match | 72.7 |
| BenCzechMark | 63.56 | Average Duel Win Score (%) | 70.5 |
| BenCzechMark - Factual Knowledge | 60.26 | Duel Win Score (%) | 70.5 |
| llmbenchmarks.io Local GPU - SimpleQA Verified | 8.5 | Correct answers (%) of 1,000 questions | 69.6 |
Interactive version: theaggregate.ai/model?slug=qwen3-8-27b-nvfp4 · How It Works · Data refreshed daily, snapshot 2026-11-05.