Qwen 3.6 35B A3B: benchmark results
Alibaba's efficiency-focused open Qwen 3.6 35B A3B MoE (3B active). Provider: Alibaba. Released 2026-04-16. Access: Open.
Unified ELO 1626 ± 1, rank #191 of 1392 rated models, from 282 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Nejumi 4 - ALT - Bias | 100 | Score (%) | 100 |
| Nejumi 4 - ALT - Toxicity | 87.52 | Score (%) | 98.4 |
| LLM Stats (MMBench-V1.1) | 92.8 | Score (%) | 97.4 |
| JudgeBench Math | 94.64 | Accuracy (%) | 97.1 |
| MERA - ruHateSpeech | 92.08 | Accuracy (%) | 96.4 |
| MERA - ruHumanEval | 76.59 | pass@1 (%) | 96.2 |
| AI for Education Visual Maths - Statistics and Probability | 71.43 | Accuracy (%) | 95.9 |
| Nejumi 4 - ALT Average | 90.2 | Score (%) | 95.9 |
| MERA | 0.79 | Total score | 95.2 |
| MERA - USE | 66.96 | Grade, normalized (%) | 94.3 |
| MERA - ruCodeEval | 70.37 | pass@1 (%) | 94.3 |
| LLM Stats (Hallusion Bench) | 69.8 | Score (%) | 94.1 |
Interactive version: theaggregate.ai/model?slug=qwen-3-6-35b-a3b · How It Works · Data refreshed daily, snapshot 2026-09-05.