Detailed comparison for LLMs
On Guardion's LLM vulnerability Benchmark, Meta Llama 4 Maverick is the more secure of the two: Llama 4 Maverick scores 27.7% and Llama 3H 405B scores 36.6% on attack success rate (ASR) (lower is better).
Llama 4 Maverick is the overall winner in this comparison!
ASR for Meta Llama 4 Maverick vs Meta Llama 3H 405B. Green marks the safer model on each metric.
Outward is better on every axis.
On Guardion's LLM vulnerability Benchmark, Meta Llama 4 Maverick is the more secure of the two: Llama 4 Maverick scores 27.7% and Llama 3H 405B scores 36.6% on attack success rate (ASR) (lower is better).
Llama 4 Maverick has a 27.7% ASR and Llama 3H 405B has a 36.6% ASR — the share of adversarial prompts that succeed across zero-shot, TAP, and Crescendo attacks. Lower is safer.
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.