Mistral Ministral 3 14B Base 2512 vs OpenAI GPT-4.1
Detailed comparison for LLMs
MistralOpenAI
Head-to-Head Overview
Ministral 3 14B Base 2512 is the overall winner in this comparison!
LLM Metric
Comparison of Attack Success Rate (ASR) metrics for Mistral Ministral 3 14B Base 2512 and OpenAI GPT-4.1 across the three attack methods.
Overall (ASR)
Ministral 3 14B Base 2512
0
GPT-4.1
34.900
TAP Attack Method (ASR)
Ministral 3 14B Base 2512
0
GPT-4.1
57.580
Crescendo Attack Method (ASR)
Ministral 3 14B Base 2512
0
GPT-4.1
44.850
Zero-Shot (ASR)
Ministral 3 14B Base 2512
0
GPT-4.1
2.290
Key Highlights
- Mistral Ministral 3 14B Base 2512 has a lower Overall (ASR).
- Mistral Ministral 3 14B Base 2512 has a lower TAP Attack Method (ASR).
- Mistral Ministral 3 14B Base 2512 has a lower Crescendo Attack Method (ASR).
- Mistral Ministral 3 14B Base 2512 has a lower Zero-Shot (ASR).
Metric Radar Chart
ministral-3-14b-base-2512
gpt-4-1