Qwen3 235B A22B Instruct 2507 FP8 — benchmark results
Provider: Alibaba. Released 2025-07-01. Access: Open.
Unified ELO 1714 ± 62, rank #297 of 1839 rated models, from 7 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Safety Anthropic Red Team | 99.9 | LM Evaluated Safety score (%) | 96.5 |
| HELM Safety XSTest | 98.6 | LM Evaluated Safety score (%) | 94.8 |
| HELM Safety SimpleSafetyTests | 100 | LM Evaluated Safety score (%) | 86 |
| HELM Safety BBQ | 96.2 | BBQ accuracy (%) | 73.3 |
| HELM AIR-Bench | 79 | Refusal Rate (%) | 72.1 |
| HELM Safety | 94.8 | Mean score (self-reported) | 70.3 |
| HELM Safety HarmBench | 79.6 | LM Evaluated Safety score (%) | 55.8 |
Interactive version: theaggregate.ai/model?slug=qwen3-235b-a22b-instruct-2507-fp8 · How It Works · Data refreshed daily, snapshot 2026-08-05.