‹ PublicAI Index
The LLM benchmark aggregator.
Phi 4
Microsoft · 15B · open weights
Strongest in Factual grounding (#4 of 104), weakest in Human preference (#263 of 354). Above par in 6 of 11 scopes. Among the models it meets almost everywhere, it finishes behind Claude 3 Opus and Claude Sonnet 5 and ahead of Claude Sonnet 4.5 and DeepSeek V3.2.
§ 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.
Safety58.2−3.2#16/3292/3
Factual grounding65.8−4.9#4/1041/1
Harm refusal59.8−1.7#11/2921/2
Jailbreak resistance62.7−2.3#18/2681/1
Toxicity avoidance57.8−1.2#34/2681/1
Secure code58.4−7.8#81/2701/1
Fairness46.5−23.5#154/2921/2
Agents47.3−19.1#172/2981/6
Tool use43.4−30.6#50/811/1
Human preference42.8−26#263/3541/1
Human preference42.8−26#263/3541/1
§ 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 Phi 4, left for the other.
§ 3 · Sources
Where the numbers come from
4 publications, 9 figures. Every one links to the page it was read from.
LMArena Text 1256
BFCL v4 28.79%
Enkrypt · Jailbreak risk 2.2%Enkrypt · Harmful content risk 3.3%Enkrypt · CBRN risk 5%Enkrypt · Toxicity risk 0.8%Enkrypt · Bias risk 83.2%Enkrypt · Insecure code risk 16%
Vectara · Factual consistency 96.3%
Badge
[](https://publicai.io/model-index/m/phi-4)