High-volume, cost-sensitive work
Choose GPT-5.6 Luna
GPT-5.6 Luna costs 3% less on a typical mix of three input tokens for every output token ($0.45 against $0.46 per million tokens).
GPT-5.4 nano vs GPT-5.6 Luna: price, context and capabilities
Choose GPT-5.6 Luna for high-volume, cost-sensitive work, writing and code generation, very long inputs, the hardest reasoning tasks, reasoning over long documents, the latest release and the most recent knowledge; on the other points the two are level or the data is missing.
In short
- GPT-5.6 Luna is 3% cheaper than GPT-5.4 nano at the headline price ($0.45 / 1M tokens vs $0.4625 / 1M tokens).
- GPT-5.6 Luna reads more per request: 1.05M tokens against 400k.
- GPT-5.6 Luna is the newer model (Jul 9, 2026, against Mar 17, 2026).
Which should you choose?
Each answer below comes straight from the published prices, specifications and independent benchmark results.
Writing and code generation
Choose GPT-5.6 Luna
GPT-5.6 Luna charges $1.20 per million output tokens, against $1.25 for GPT-5.4 nano. Output dominates the bill when answers are long.
Very long inputs
Choose GPT-5.6 Luna
GPT-5.6 Luna reads up to 1.05M tokens per request, against 400k for GPT-5.4 nano.
Hardest reasoning tasks
Choose GPT-5.6 Luna
GPT-5.6 Luna scores 37.3 on the Artificial Analysis Intelligence Index, against 20.7 for GPT-5.4 nano.
Reasoning over long documents
Choose GPT-5.6 Luna
GPT-5.6 Luna scores 83.7% on Long-context reasoning (AA-LCR), against 76.7% for GPT-5.4 nano.
Latest release
Choose GPT-5.6 Luna
GPT-5.6 Luna is the newer release (Jul 9, 2026, against Mar 17, 2026 for GPT-5.4 nano).
Most recent knowledge
Choose GPT-5.6 Luna
GPT-5.6 Luna was trained on data up to February 2026, against August 2025 for GPT-5.4 nano.
Monthly cost at three volumes
| Volume | GPT-5.4 nano | GPT-5.6 Luna | Difference |
|---|---|---|---|
| Prototype: 1M input, 250K output tokens | $0.51 | $0.50 | GPT-5.6 Luna, $0.01 less |
| Production: 20M input, 5M output tokens | $10.25 | $10.00 | GPT-5.6 Luna, $0.25 less |
| Scale: 500M input, 100M output tokens | $225.00 | $220.00 | GPT-5.6 Luna, $5.00 less |
Standard rates, without caching or batch discounts. Volumes are per month.
Specifications side by side
| GPT-5.4 nano | GPT-5.6 Luna | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Category | Language models | Language models |
| Status | Deprecated | Generally available |
| Headline price | $0.4625 / 1M tokens | $0.45 / 1M tokens |
| Input | $0.20 / 1M tokens | $0.20 / 1M tokens |
| Output | $1.25 / 1M tokens | $1.20 / 1M tokens |
| Cached input | $0.02 / 1M tokens | $0.02 / 1M tokens |
| Context window | 400k | 1.05M |
| Max output | 128k | 128k |
| Released | Mar 17, 2026 | Jul 9, 2026 |
| Knowledge cutoff | August 2025 | February 2026 |
| Licence | Proprietary | Proprietary |
| Capabilities | Vision, Reasoning, Tool calling, Structured output, Prompt caching, Batch API | Vision, Reasoning, Tool calling, Structured output, Prompt caching, Batch API |
| API model id | gpt-5.4-nano | gpt-5.6-luna |
Benchmarks side by side
Leader in bold.
| Benchmark | GPT-5.4 nano | GPT-5.6 Luna |
|---|---|---|
| Intelligence Index | 20.7 | 37.3 |
| SciCode | 47.2% | 53.6% |
| GPQA Diamond | 81.7% | 91.1% |
| Humanity's Last Exam | 28.3% | 39.5% |
| MMMU-Pro (visual reasoning) | 65.4% | 78.6% |
| Long-context reasoning (AA-LCR) | 76.7% | 83.7% |
| Terminal-Bench Hard | 42.4% | No data |
| IFBench (instruction following) | 75.9% | No data |
| τ²-Bench (agentic tool use) | 76% | No data |
Independent evaluations by Artificial Analysis, run the same way for every model, at the reasoning setting shown. Percentages unless marked as an index.
Frequently asked questions
Should I choose GPT-5.4 nano or GPT-5.6 Luna?
Choose GPT-5.6 Luna for high-volume, cost-sensitive work, writing and code generation, very long inputs, the hardest reasoning tasks, reasoning over long documents, the latest release and the most recent knowledge; on the other points the two are level or the data is missing.
Which is cheaper, GPT-5.4 nano or GPT-5.6 Luna?
GPT-5.6 Luna is 3% cheaper than GPT-5.4 nano at the headline price ($0.45 / 1M tokens vs $0.4625 / 1M tokens).
Which has the larger context window, GPT-5.4 nano or GPT-5.6 Luna?
GPT-5.6 Luna reads more per request: 1.05M tokens against 400k.
What is the difference between GPT-5.4 nano and GPT-5.6 Luna?
GPT-5.4 nano is made by OpenAI and GPT-5.6 Luna by OpenAI. GPT-5.6 Luna is 3% cheaper than GPT-5.4 nano at the headline price ($0.45 / 1M tokens vs $0.4625 / 1M tokens). GPT-5.6 Luna reads more per request: 1.05M tokens against 400k. GPT-5.6 Luna is the newer model (Jul 9, 2026, against Mar 17, 2026).
How much do GPT-5.4 nano and GPT-5.6 Luna cost per month?
At 20 million input and 5 million output tokens a month, GPT-5.4 nano costs $10.25 and GPT-5.6 Luna costs $10.00, at standard rates.
Which is newer, GPT-5.4 nano or GPT-5.6 Luna?
GPT-5.6 Luna was released on Jul 9, 2026 and GPT-5.4 nano on Mar 17, 2026.