Qwen 3 235B A22B 2507 FP8 (Thinking): benchmark results
Provider: Alibaba. Released 2025-07-25. Access: Open.
Unified ELO 1609 ± 1, rank #388 of 1935 rated models, from 164 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval German | 68.41 | Average Score (%) | 100 |
| EuroEval Italian | 73.12 | Average Score (%) | 100 |
| EuroEval Lithuanian | 66.49 | Average Score (%) | 100 |
| EuroEval Lithuanian Knowledge | 91.04 | Knowledge Average Score (%) | 100 |
| EuroEval Portuguese | 73.86 | Average Score (%) | 100 |
| EuroEval Spanish | 68.78 | Average Score (%) | 100 |
| EuroEval French | 73.27 | Average Score (%) | 99.7 |
| EuroEval German NLU - GermEval | 78.33 | Named entity recognition Score (%) | 99.7 |
| EuroEval Slovak NLU - UNER SK | 73.22 | Named entity recognition Score (%) | 99.5 |
| EuroEval Finnish NLU - Turku NER FI | 79.42 | Named entity recognition Score (%) | 99.2 |
| EuroEval English | 77.99 | Average Score (%) | 99.1 |
| EuroEval Italian NLU - MultiNERD IT | 88.01 | Named entity recognition Score (%) | 99.1 |
Interactive version: theaggregate.ai/model?slug=qwen-3-235b-a22b-2507-fp8-thinking · How It Works · Data refreshed daily, snapshot 2026-09-25.