Alibaba Qwen 7-72BB vs DeepSeek DeepSeek-V3.2-Speciale

Detailed comparison for LLMs

AlibabaDeepSeek

On Guardion's LLM vulnerability Benchmark, Alibaba Qwen 7-72BB is the more secure of the two: Qwen 7-72BB scores 36.5% and DeepSeek-V3.2-Speciale scores 45.0% on attack success rate (ASR) (lower is better). One or both scores are estimated from public safety evaluations pending a Guardion benchmark run.

Head-to-Head Overview

Qwen 7-72BB is the overall winner in this comparison!

Attack Success Rate (lower is safer)

ASR for Alibaba Qwen 7-72BB vs DeepSeek DeepSeek-V3.2-Speciale. Green marks the safer model on each metric. Only the overall score is available for estimated models.

Overall (ASR)

Qwen 7-72BB
36.5%
DeepSeek-V3.2-Speciale
45.0%

TAP Attack Method (ASR)

Qwen 7-72BB
DeepSeek-V3.2-Speciale
100.0%

Crescendo Attack Method (ASR)

Qwen 7-72BB
DeepSeek-V3.2-Speciale
100.0%

Zero-Shot (ASR)

Qwen 7-72BB
9.5%
DeepSeek-V3.2-Speciale
100.0%

Key Highlights

  • Alibaba Qwen 7-72BB has a lower Overall (ASR).
  • Alibaba Qwen 7-72BB has a lower Zero-Shot (ASR).

Security Profile

Outward is better on every axis.

OverallTAPCrescendoZero-Shot
Qwen 7-72BB
DeepSeek-V3.2-Speciale
Full security profile
Alibaba Qwen 7-72BB
Full security profile
DeepSeek DeepSeek-V3.2-Speciale

Frequently asked questions

Is Alibaba Qwen 7-72BB or DeepSeek DeepSeek-V3.2-Speciale more secure?

On Guardion's LLM vulnerability Benchmark, Alibaba Qwen 7-72BB is the more secure of the two: Qwen 7-72BB scores 36.5% and DeepSeek-V3.2-Speciale scores 45.0% on attack success rate (ASR) (lower is better). One or both scores are estimated from public safety evaluations pending a Guardion benchmark run.

What is the attack success rate (ASR) of Qwen 7-72BB vs DeepSeek-V3.2-Speciale?

Qwen 7-72BB has a 36.5% ASR and DeepSeek-V3.2-Speciale has a 45.0% ASR — the share of adversarial prompts that succeed across zero-shot, TAP, and Crescendo attacks. Lower is safer.

How were Qwen 7-72BB and DeepSeek-V3.2-Speciale tested?

Both were red-teamed with the HarmBench framework across zero-shot, TAP (Tree of Attacks with Pruning), and Crescendo multi-turn attacks, scored by Attack Success Rate.

Related Comparisons