Phi-4 vs GPT-5 mini

Phi-4 is cheaper than GPT-5 mini by about 90% on a blended (75% input / 25% output) basis. Phi-4 charges $0.0560 / $0.11 per 1M input/output tokens; GPT-5 mini charges $0.25 / $2.00.

FieldPhi-4GPT-5 mini
Input $/Mtok$0.0560$0.25
Output $/Mtok$0.11$2.00
Cache read $/Mtok–$0.0250
Cache write $/Mtok––
Input $/Mcharn/a0.0704
Output $/Mcharn/a0.56
Context window16384400000
Tokenizer–o200k
Confidence?apiapi
Sourceopenrouteropenrouter

Confidence: official = vendor pricing page · api = OpenRouter per-host · aggregator = community list, may lag · unreviewed = automatic extraction awaiting review. When sources disagree the most authoritative wins. Details: /methodology#confidence

Price history

No price changes recorded yet for either model.

Same workload, monthly cost

Illustrative monthly cost at 10M tokens/month, split per workload preset. Not a substitute for the calculator, which uses your own volume, cache and batch assumptions.

WorkloadPhi-4GPT-5 mini
Coding agent$0.59$3.38
RAG chat$0.64$5.13
Summarisation (batch)$0.62$4.25
Classification$0.57$2.85
Chat assistant$0.78$9.50
Content generation$1.01$16.50

Batch pricing not published for Phi-4; standard rate used for the batch share of a workload.

Batch pricing not published for GPT-5 mini; standard rate used for the batch share of a workload.

Cite this page

Phi-4 vs GPT-5 mini: $0.0560/$0.11 vs $0.25/$2.00 per 1M tokens as of 2026-11-19. Token Barometer by Dejan Strbad, https://tokenbarometer.com, data CC BY 4.0. https://tokenbarometer.com/compare/microsoft_phi-4-vs-openai_gpt-5-mini

Phi-4 page | GPT-5 mini page | Open Phi-4 in calculator | Open GPT-5 mini in calculator