Comparison

GPT-4o vs GPT-5.4 mini

Per input token, GPT-5.4 mini runs 70% cheaper than GPT-4o. GPT-5.4 mini is 55% cheaper on output tokens. GPT-5.4 mini has the larger context window than GPT-4o.

FieldGPT-4o
Legacy
GPT-5.4 mini
ProviderOpenAIOpenAI
Context window128K400K
Max output128K
Input / 1M$2.50$0.7570% cheaper
Output / 1M$10.00$4.5055% cheaper
Cached input / 1M$1.25$0.07594% cheaper
Batch discount50%
Release date2026-03-17

Worked cost examples

Estimated monthly cost on three common usage shapes

WorkloadGPT-4oGPT-5.4 mini

Chatbot

800 input / 200 output tokens per request, 100,000 requests/month

$400.00$150.00cheaper

RAG pipeline

6,000 input / 700 output tokens per request, 30,000 requests/month

$660.00$229.50cheaper

Batch summarization

20,000 input / 1,500 output tokens per request, 5,000 requests/month

$325.00$108.75cheaper

GPT-4o vs GPT-5.4 mini FAQ

Is GPT-4o cheaper than GPT-5.4 mini?
GPT-4o costs $2.50 per 1M input tokens versus $0.75 for GPT-5.4 mini — GPT-5.4 mini is 233% cheaper per input token.
Which is cheaper for output tokens, GPT-4o or GPT-5.4 mini?
GPT-4o charges $10.00 per 1M output tokens; GPT-5.4 mini charges $4.50 per 1M output tokens.
Which has the larger context window, GPT-4o or GPT-5.4 mini?
GPT-5.4 mini does, with a 400,000-token context window.
Do GPT-4o and GPT-5.4 mini offer cache pricing?
GPT-4o discounts cached input to $1.25 per 1M tokens. GPT-5.4 mini discounts cached input to $0.075 per 1M tokens.

Looking for another matchup? See all comparisons.