Text model
gpt-5.4-mini pricing
OpenAI charges $0.75 per million input tokens and $4.50 per million output tokens for gpt-5.4-mini, and $0.075 for cached input. Below: every rate that applies, what common workloads cost, and cheaper options.
- Input
- $0.75/ 1M
- Output
- $4.50/ 1M
- Cached input
- $0.075/ 1M
- Context
- 272Ktokens
All rates
| Tier | Input | Output | Cache read | Cache write |
|---|---|---|---|---|
| Standard | $0.75 | $4.50 | $0.075 | — |
| Batch | $0.375 | $2.25 | $0.0375 | — |
| Priority | $1.50 | $9.00 | $0.15 | — |
| Flex | $0.375 | $2.25 | $0.0375 | — |
USD per million tokens. Long-context tiers apply when the whole prompt, cached tokens included, passes the threshold.
Work it out
What common workloads cost
| Workload | Per call | Cost per call | Calls a month | Per month |
|---|---|---|---|---|
| Chat reply | 1,500 input and 400 output tokens | $0.00293 | 100,000 | $292.50 |
| RAG answer, cached context | 2,000 fresh + 4,000 cached input, 500 output | $0.00405 | 100,000 | $405.00 |
| Agent step | 4,000 fresh + 8,000 cached input, 800 output | $0.0072 | 250,000 | $1,800.00 |
| Batch summary | 3,000 input and 300 output tokens via the Batch API | $0.0018 | 1,000,000 | $1,800.00 |
Planning a mix of models? Simulate a workload with gpt-5.4-mini, alongside anything else you run.
Cheaper options
Current models that cost at least 15% less for a typical chat reply. Cheaper isn't always equivalent: test quality on your own prompts, then use veqen diff to see the impact at your volume.
| Model | Chat reply | Saving |
|---|---|---|
| openai/gpt-5-mini | $0.00118 | 60% |
| openai/gpt-5.6-luna | $0.00078 | 73% |
| bedrock/amazon.nova-pro-v1:0 | $0.00248 | 15% |
| databricks/databricks-gemini-3-flash | $0.00244 | 17% |
Price history
No price changes since Veqen started tracking this model.
Also billed under: gpt-5.4-mini-2026-03-17.
Compare
Last verified with price book 2026-10-22. Sources: OpenAI pricing and LiteLLM's MIT-licensed price file. Track every call to this model, per customer or project: install the SDK.