OpenAI GPT-4o vs Google Gemini 2.0 Flash Thinking
Detailed comparison for LLMs
On Guardion's LLM vulnerability Benchmark, Google Gemini 2.0 Flash Thinking is the more secure of the two: GPT-4o scores 22.8% and Gemini 2.0 Flash Thinking scores 21.5% 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
Gemini 2.0 Flash Thinking is the overall winner in this comparison!
Attack Success Rate (lower is safer)
ASR for OpenAI GPT-4o vs Google Gemini 2.0 Flash Thinking. Green marks the safer model on each metric. Only the overall score is available for estimated models.
Overall (ASR)
TAP Attack Method (ASR)
Crescendo Attack Method (ASR)
Zero-Shot (ASR)
Key Highlights
- Google Gemini 2.0 Flash Thinking has a lower Overall (ASR).
- OpenAI GPT-4o has a lower TAP Attack Method (ASR).
- OpenAI GPT-4o has a lower Crescendo Attack Method (ASR).
- OpenAI GPT-4o has a lower Zero-Shot (ASR).
Security Profile
Outward is better on every axis.
Frequently asked questions
Is OpenAI GPT-4o or Google Gemini 2.0 Flash Thinking more secure?
On Guardion's LLM vulnerability Benchmark, Google Gemini 2.0 Flash Thinking is the more secure of the two: GPT-4o scores 22.8% and Gemini 2.0 Flash Thinking scores 21.5% 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 GPT-4o vs Gemini 2.0 Flash Thinking?
GPT-4o has a 22.8% ASR and Gemini 2.0 Flash Thinking has a 21.5% ASR — the share of adversarial prompts that succeed across zero-shot, TAP, and Crescendo attacks. Lower is safer.
How were GPT-4o and Gemini 2.0 Flash Thinking 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.