OpenAI Pricing Calculator (GPT-5, GPT-4o, gpt-oss)
| Model | Input / 1M | Output / 1M | Context |
|---|---|---|---|
GPT-5.5 (xhigh) OpenAI | $1.25 | $10.00 | 922K |
GPT-5.5 (high) OpenAI | $1.25 | $10.00 | 922K |
GPT-5.5 (medium) OpenAI | $1.25 | $10.00 | 922K |
GPT-5.5 (low) OpenAI | $1.25 | $10.00 | 922K |
GPT-5.3 Codex (xhigh) OpenAI | $1.25 | $10.00 | 400K |
GPT-5.4 mini (xhigh) OpenAI | $0.25 | $2.00 | 400K |
GPT-5.4 nano (xhigh) OpenAI | $0.05 | $0.40 | 400K |
o3 OpenAI | $2.00 | $8.00 | 200K |
GPT-4o OpenAI | $2.50 | $10.00 | 128K |
GPT-4o mini OpenAI | $0.15 | $0.60 | 128K |
gpt-oss 120B OpenAI | $0.039 | $0.10 | 131K |
gpt-oss 20B OpenAI | $0.03 | $0.14 | 131K |
Which OpenAI model should you use?
GPT-5.5 is the new default for production — smarter than GPT-4o at half the input cost. GPT-5.4 mini ($0.25/1M input) handles 80% of chat and RAG workloads at a fraction of the price. GPT-5.4 nano ($0.05/1M) is the cheapest first-class option; use it for classification, routing, and high-volume tool calls.
GPT-4o vs GPT-5 pricing
GPT-4o is still listed but is almost never the right choice in 2026 — GPT-5.5 is cheaper on input, matches on output, and scores higher on every public benchmark. Migration is a one-line model-name change.
How to lower your OpenAI bill
Three levers, ordered by impact: (1) Enable prompt caching for any system prompt over 1,024 tokens — 50% off the cached portion. (2) Use the Batch API for non-realtime workloads — another 50% off. (3) Tier down: a lot of GPT-5.5 traffic can move to GPT-5.4 mini with no measurable quality drop.