GPT-5.4 nano vs GPT-6 Luna
OpenAI has replaced GPT-5.4 nano with GPT-6 Luna. There is no quality verdict yet: the AIGROW small-business benchmark has not run GPT-6 Luna. On price, GPT-6 Luna costs less: $0.10 in and $0.50 out per million tokens, against $0.20 in and $1.25 out for GPT-5.4 nano.
Run GPT-6 Luna up to 75% below list price. $25 a month brings $110 of API credit and cancels any time; $10 once buys $40. For long documents, GPT-6 Luna reads up to 1.05M tokens at once and GPT-5.4 nano up to 400K. US dollars, standard rate, shortest context band. Price only: the benchmark has not run both. Prices checked 18 November 2026.
- GPT-5.4 nano
- $1.10
- GPT-6 Luna
- $0.50
- GPT-5.4 nano, per M tokens
- $0.20 / $1.25
- GPT-6 Luna, per M tokens
- $0.10 / $0.50
Migrating from deprecated GPT-5.4 nano to OpenAI's recommended GPT-6 Luna cuts input by half and output by 60%. 1,000 support replies cost $1.10 on GPT-5.4 nano and $0.50 on GPT-6 Luna, so GPT-5.4 nano costs 2.2 times as much. GPT-5.4 nano was released on 17 Mar 2026, GPT-6 Luna was released on 22 Sep 2026.
What each one costs
The API price per million tokens and the cheapest app plan with each model, from the makers’ own pages. Each name opens that model’s full price page.
- GPT-5.4 nano
- On the OpenAI API, GPT-5.4 nano costs $0.20 per million input tokens and $1.25 per million output tokens.
- GPT-6 Luna
- On the OpenAI API, GPT-6 Luna costs $0.10 per million input tokens and $0.50 per million output tokens. In ChatGPT, it comes with the free plan, and the cheapest paid plan with it is Go at $8/mo. Plan prices read 14 October 2026. AIGROW API credit runs GPT-6 Luna at $0.025 / $0.125 per million input / output tokens, up to 75% below list price. See API credit
$25 a month brings $110 of API credit and cancels any time; $10 once buys $40.
Side by side
Each maker’s published rate for its standard tier. Cached input is what a repeated prompt prefix costs once the provider has stored it.
| Fact | GPT-5.4 nano | GPT-6 Luna |
|---|---|---|
| Maker | OpenAI | OpenAI |
| Tier | Replaced by GPT-6 Luna | Small and fast |
| Input, per M tokens | $0.20 | $0.10 |
| Output, per M tokens | $1.25 | $0.50 |
| Cached input | $0.02 | $0.01 |
| Context window | 400K | 1.05M |
| Released | 17 Mar 2026 | 22 Sep 2026 |
| Weights | Closed weights | Closed weights |
What a job costs on each
1,000 runs of each job at the two list prices. The gap moves from job to job because the models price input, output and cached context differently.
| Job | GPT-5.4 nano | GPT-6 Luna | Cheaper |
|---|---|---|---|
| Support reply3,000 in, 400 out. | $1.10 | $0.50 | GPT-6 Luna, 55% less |
| Document summary25,000 in, 1,000 out. | $6.25 | $3.00 | GPT-6 Luna, 52% less |
| Agent task80,000 in (70,000 cached), 3,000 out. | $7.15 | $3.20 | GPT-6 Luna, 55% less |
The token counts are ours, chosen to look like real work; the prices are the makers’. A model with no published cache price bills the agent task’s repeated context at its full input rate.
The small print
What each base rate leaves out, from the makers’ own pages: surcharges, discounts, promotions and cache pricing.
- GPT-5.4 nano
- Deprecated on 1 October 2026: OpenAI shuts it down on 1 April 2027 and recommends GPT-6 Luna instead. Batch 50% off. OpenAI’s pricing page
- GPT-6 Luna
- Requests above 272K input tokens bill 2x input and 1.5x output ($0.20 / $0.75, cached $0.02). Cache write $0.125. Batch and Flex 50% off, Fast mode 2x. Data residency adds 10%. OpenAI’s pricing page
Other comparisons
Every head to head on the list that involves GPT-5.4 nano or GPT-6 Luna.
- Claude Haiku 5.5 vs GPT-6 Luna
- GPT-6 Luna vs Gemini 3.5 Flash-Lite
- GPT-6.1 Terra vs GPT-6 Luna
- Muse Ember vs GPT-6 Luna
- DeepSeek V4.1 Flash vs GPT-6 Luna
- GPT-6 Luna vs Gemini 3.8 Flash
- GPT-6 Luna vs Claude Opus 5.5
Comparisons with an earlier generation
Run GPT-6 Luna up to 75% below list price
AIGROW API credit covers GPT-6 Luna at published rates never above the list prices here, and $10 buys $40 of credit.
Prices checked 18 November 2026