100,000 GPT-5.4 mini tokens cost $0.075 as input or $0.45 as output. Most workloads land in between — a 70/30 input/output split costs about $0.1875. Verified Aug 16, 2026
Interactive calculator
100,000 in · 0 out
Input price
$0.75 / 1M tokens
Output price
$4.50 / 1M tokens
Uncached pricing. Add caching and monthly projections in the full calculator.
What 100,000 tokens buys in practice
$0.15
80% input / 20% output — short replies over conversation history.
$0.1125
90% input / 10% output — long retrieved context, grounded answers.
$0.0938
95% input / 5% output — full documents in, structured extraction out.
Same 100,000 tokens on other models
70/30 input/output split, uncached.
FAQ
100,000 input tokens cost $0.075; 100,000 output tokens cost $0.45. A typical 70/30 input/output mix costs about $0.1875. Rates: $0.75/M input, $4.50/M output.
A chat-style request of ~4,800 tokens (4,000 in + 800 out) means 100,000 tokens is roughly 21 requests. For a RAG workload with ~8,500 tokens per request, about 12 requests.
Yes — cached input costs $0.075/M instead of $0.75/M, a 90% discount on repeated context.