CodeSmart
← All models

GPT-5.4 mini

OpenAI

AA benchmark indicesIndices from Artificial Analysis. Intelligence score = 50% agentic + 40% coding + 10% speed.Read methodology →

AAIntelligence scoreVerified
55.5/100

Intelligence score composite

AAIntelligence
—
AACodingVerified
56.1/100
AAAgenticVerified
56.1/100
AA quality trend
Not enough history

Best-value plans for this model

Best low-cost
No plan offers this model in this band.
Best mid-tier
No plan offers this model in this band.
Best high-end
No plan offers this model in this band.

Plans offering GPT-5.4 mini

No plan in the dataset lists this exact model id.