Qwen3.6-35B-A3B-NVFP4: benchmark results
Provider: Alibaba. Released 2026-05-28. Access: Open.
Aggregate Rating 1614 ± 38, rank #887 of 10916 rated models, from 14 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| llmbenchmarks.io Local GPU - SimpleQA Verified | 19.4 | Correct answers (%) of 1,000 questions | 100 |
| BenCzechMark - Reading Comprehension | 75.64 | Duel Win Score (%) | 92.3 |
| BenCzechMark - Czech Language Understanding | 68.16 | Duel Win Score (%) | 84.6 |
| llmbenchmarks.io Local GPU - IFBench | 56.1 | Instruction-level strict accuracy (%) | 82.6 |
| llmbenchmarks.io Local GPU - Chess Puzzles | 6 | Puzzles solved (%), exact match | 81.8 |
| BenCzechMark - NER | 78.85 | Duel Win Score (%) | 80.8 |
| BenCzechMark | 65.56 | Average Duel Win Score (%) | 75.6 |
| BenCzechMark - Factual Knowledge | 61.17 | Duel Win Score (%) | 73.1 |
| BenCzechMark - Czech Math Reasoning | 60.58 | Duel Win Score (%) | 71.8 |
| llmbenchmarks.io Local GPU - IFEval | 86.69 | Prompt-level strict accuracy (%) | 65.2 |
| llmbenchmarks.io Local GPU - SciCode | 33.56 | Subproblems solved (%), pass@1 averaged over 3 epochs | 63.6 |
| BenCzechMark - NLI | 54.57 | Duel Win Score (%) | 60.3 |
Interactive version: theaggregate.ai/model?slug=qwen3-6-35b-a3b-nvfp4 · How It Works · Data refreshed daily, snapshot 2026-11-08.