‹ PublicAI Index
The LLM benchmark aggregator.
O1 Mini
OpenAI
Strongest in Jailbreak resistance (#24 of 272), weakest in Human preference (#180 of 342). Above par in 11 of 15 scopes. Among the models it meets almost everywhere, it finishes behind GPT-5 and O3 and ahead of Claude Haiku 4.5 and GLM-4.5.
§ 1 · Profile
What it is good at
Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.
Safety56.6−4.9#29/3372/3
Jailbreak resistance61.8−3.1#24/2721/1
Safe-prompt compliance56.4−4#27/821/1
Harm refusal56.9−4.7#35/3002/2
Secure code60.2−6.3#67/2741/1
Fairness53−17#83/3002/2
Toxicity avoidance55.1−3.9#107/2721/1
Professional51.4−10.2#67/1681/1
Medical53.6−13.2#54/1401/1
Coding47.5−21.6#90/1651/5
Agentic coding47−21.7#86/1571/4
Reasoning40−26.9#153/1782/4
Reasoning35.5−33.6#133/1402/3
Human preference50.9−16.6#180/3421/1
Human preference50.9−16.6#180/3421/1
§ 2 · Head to head
What it beats, and what beats it
The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for O1 Mini, left for the other.
§ 3 · Sources
Where the numbers come from
7 publications, 16 figures. Every one links to the page it was read from.
LMArena Text 1337
ARC-AGI-2 0.8%
Aider polyglot 32.9%
SimpleBench 18.1%
Vals · MedQA 90.22%
HELM Safety · HarmBench 88.5%HELM Safety · SimpleSafetyTests 97%HELM Safety · Anthropic Red Team 98.3%HELM Safety · BBQ 96.9%HELM Safety · XSTest 97%
Enkrypt · Jailbreak risk 2.8%Enkrypt · Harmful content risk 7.2%Enkrypt · CBRN risk 7.3%Enkrypt · Toxicity risk 2.7%Enkrypt · Bias risk 80.9%Enkrypt · Insecure code risk 12.9%
Badge
[](https://publicai.io/model-index/m/o1-mini)