Anthropic Claude 3.5 Sonnet v2 vs Cohere Command R 7B
Detailed comparison for LLMs
On Guardion's LLM vulnerability Benchmark, Anthropic Claude 3.5 Sonnet v2 is the more secure of the two: Claude 3.5 Sonnet v2 scores 4.4% and Command R 7B scores 25.8% on attack success rate (ASR) (lower is better).
Head-to-Head Overview
Claude 3.5 Sonnet v2 is the overall winner in this comparison!
Attack Success Rate (lower is safer)
ASR for Anthropic Claude 3.5 Sonnet v2 vs Cohere Command R 7B. Green marks the safer model on each metric.
Overall (ASR)
TAP Attack Method (ASR)
Crescendo Attack Method (ASR)
Zero-Shot (ASR)
Key Highlights
- Anthropic Claude 3.5 Sonnet v2 has a lower Overall (ASR).
- Anthropic Claude 3.5 Sonnet v2 has a lower TAP Attack Method (ASR).
- Anthropic Claude 3.5 Sonnet v2 has a lower Crescendo Attack Method (ASR).
Security Profile
Outward is better on every axis.
Frequently asked questions
Is Anthropic Claude 3.5 Sonnet v2 or Cohere Command R 7B more secure?
On Guardion's LLM vulnerability Benchmark, Anthropic Claude 3.5 Sonnet v2 is the more secure of the two: Claude 3.5 Sonnet v2 scores 4.4% and Command R 7B scores 25.8% on attack success rate (ASR) (lower is better).
What is the attack success rate (ASR) of Claude 3.5 Sonnet v2 vs Command R 7B?
Claude 3.5 Sonnet v2 has a 4.4% ASR and Command R 7B has a 25.8% ASR — the share of adversarial prompts that succeed across zero-shot, TAP, and Crescendo attacks. Lower is safer.
How were Claude 3.5 Sonnet v2 and Command R 7B 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.