GPT-5.4 Mini
by OpenAI · openai/gpt-5.4-mini · #217 cheapest of 325 paid models
Prices updated Jul 22, 2026, 1:31 AM UTC · refreshed hourly
GPT-5.4 mini brings the core capabilities of GPT-5.4 to a faster, more efficient model optimized for high-throughput workloads. It supports text and image inputs with strong performance across reasoning, coding,...
Compare vs:Claude Sonnet 5Claude Opus 4.6Claude Haiku 4.5Gemini 3.6 Flash
Pricing
Input
$0.75
per 1M tokens
Output
$4.50
per 1M tokens
Blended (3:1)
$1.69
3 input : 1 output
Cache read
$0.075
per 1M cached tokens
Cache write
—
per 1M tokens
Cache & batch economics
Effective prices for GPT-5.4 Mini with prompt caching.
| Cache hit rate | Effective blended $/1M | Example request* | vs no cache |
|---|---|---|---|
| 0% | $1.69 | $0.081 | — |
| 50% | $1.43 | $0.0473 | −42% |
| 90% | $1.23 | $0.0203 | −75% |
*Example: 100K sticky context + 2K new input + 1K output. Batch discount is an approximate 50% off list rates for this provider’s batch API. Confirm on the vendor’s pricing page.
What a request costs
| Workload | Input tokens | Output tokens | Cost / request | Cost / 1K requests |
|---|---|---|---|---|
| Short chat message | 500 | 300 | $0.001725 | $1.72 |
| Document summary | 8,000 | 1,000 | $0.0105 | $10.50 |
| Codebase question (RAG) | 30,000 | 2,000 | $0.0315 | $31.50 |
| Long-context analysis | 150,000 | 5,000 | $0.135 | $135.00 |
Performance
Intelligence Index
40
AA composite quality
Coding Index
56.1
Math Index
—
Agentic Index
30.2
Output speed
180 tok/s
median
Time to first token
10.39s
median
Time to first answer
10.39s
after reasoning tokens
Artificial Analysis benchmarks
| Benchmark | Score |
|---|---|
| GPQA Diamond | 87.5% |
| Humanity's Last Exam | 26.6% |
| SciCode | 49.9% |
| IFBench | 73.3% |
| AA-LCR (Long Context Reasoning) | 69.3% |
| Terminal-Bench Hard | 52.3% |
| Terminal-Bench 2.1 | 59.2% |
| τ²-Bench (Telecom) | 83.3% |
| τ³-Bench Banking | 21.4% |
Specs
Context window
400K
tokens
Max output
128K
tokens
Input modalities
file, image, text
Tokenizer
GPT