GPT-4o mini
Fast / cheapLast verified: June 2026· OpenAI GPT-4o mini pricing
Low-cost small model with a 128K context window. Confirm current pricing with OpenAI.
- Input / 1M
- $0.15
- Output / 1M
- $0.60
- Context
- 128K tokens
- Provider
- OpenAI
Pricing verified June 2026. Prices change frequently. Always confirm against the provider’s official pricing page before relying on these figures for budgeting. Official pricing →
When to choose GPT-4o mini
The cheapest row in this table: $0.15 per million in and $0.60 out, which puts a month of 50M input and 10M output tokens at roughly $14 all in. That number is the floor every routing exercise is measured against — the target you are trying to move low-stakes traffic towards. It keeps the same 128K window as GPT-4o, so the ceiling you meet first is capability rather than context. Use it for classification, extraction, tagging and first-pass drafts, and gate the escalation boundary with evals.
GPT-4o mini sits close enough to GPT-4.1 mini on price that the two are usually decided together. That comparison is written once, on the models table, rather than argued twice from opposite sides.
What GPT-4o mini costs in practice
Everything about this row is the floor. The 50M/10M month is $13.50. A request that fills the whole 128K window costs under two cents of input. The next row up, Llama 3.3 70B, is only about 14% dearer blended, so the interesting comparison is never with other cheap models — it is with whatever you are running today. If a route currently sits on a mid-tier model at $180 a month, the question worth putting to an eval suite is what actually breaks when it moves here.
| Scenario | Input / mo | Output / mo | Est. cost / mo |
|---|---|---|---|
| Prototype | 2M | 0.5M | $1 |
| Growing product | 50M | 10M | $14 |
| At scale | 500M | 100M | $135 |
Plug in your own numbers with the cost calculator.
Where GPT-4o mini fits in a routing setup
This is where routed traffic lands rather than where it starts. Nothing on this table is cheaper, so the routing work all points upward: which requests fail here and need GPT-4o at $225 on the reference month against $13.50. The nearest alternative at this price is Llama 3.3 70B at $16, which swaps a hosted API for an open-weight one. The gate matters more than the model, because a cheap answer that has to be redone twice was never cheap.
Frequently asked questions
How much does GPT-4o mini cost?
GPT-4o mini bills $0.15 per million input tokens and $0.60 per million output tokens — $0.26 per million blended at a three-to-one input:output split. A workload of 50M input and 10M output tokens per month works out at about $14. Confirm current pricing with the provider.
What is GPT-4o mini's context window?
GPT-4o mini has a 128K-token context window. One request that fills the whole window costs about $0.02 at the input rate, before the reply is billed.
Where does GPT-4o mini sit on price?
At $0.26 per million blended — three input tokens for every output token — GPT-4o mini is nothing on this page is cheaper, and 18 models are dearer. Output bills at 4x the input rate.
Related concepts: context window, model routing, prompt caching, evals.