Qwen 3 30B A3B — benchmark results
Alibaba's open 30B A3B MoE (3.3B active) with switchable thinking/non-thinking modes, efficiency-focused. Provider: Alibaba. Released 2025-04-28. Access: Open.
Unified ELO 1569 ± 7, rank #568 of 1776 rated models, from 456 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Latvian NLU - Latvian Twitter Sentiment | 53.11 | Sentiment classification Score (%) | 97.8 |
| Open-R1 Eval Leaderboard | 72.94 | Average Accuracy (%) | 97.2 |
| EuroEval Italian NLU - MultiNERD IT | 84.97 | Named entity recognition Score (%) | 97 |
| Evals for Every Language - Language la | 100 | Average Score (%) | 96.3 |
| Agent-ValueBench | 7.8 | Authority (self-reported) | 96.2 |
| Fibble5 Arena | 30 | Win Rate (%) | 95.6 |
| EuroEval German NLU - GermEval | 73.17 | Named entity recognition Score (%) | 95 |
| EuroEval Lithuanian | 58.06 | Average Score (%) | 94.3 |
| EuroEval Polish Common Sense Reasoning | 60.73 | Common Sense Reasoning Average Score (%) | 93 |
| EuroEval Polish | 60.61 | Average Score (%) | 92.7 |
| EuroEval Spanish NLU - CoNLL ES | 77.85 | Named entity recognition Score (%) | 92.7 |
| EuroEval Italian | 64.66 | Average Score (%) | 92.5 |
Interactive version: theaggregate.ai/model?slug=qwen-3-30b-a3b · How the rankings work · Data refreshed daily, snapshot 2026-07-22.