Guides / gpt-5.6 luna pricing
GPT-5.6 Luna pricing at high call volume
GPT-5.6 Luna pricing is $0.20 per million input tokens and $1.20 per million output tokens, the cheapest OpenAI tier and fractionally under GPT-5.4 nano on output. At this rate call volume decides your bill rather than the rate. Culpa, a local-first LLM cost, margin, and forecast ledger, meters every call so a cheap tier can't hide behind its price.
Why this happens
Luna and GPT-5.4 nano share an identical $0.20 input rate and differ by five cents on output, $1.20 against $1.25. At that level the choice between them isn't a cost decision at all. What matters is what happens after you pick one. A rate this low stops feeling worth instrumenting, so retry caps loosen, prompts stop getting reviewed, and the model handling your largest call count becomes the least watched line on the bill.
What this usually looks like
- A Luna-powered feature appears in your top three by spend, which nobody predicted.
- Luna traffic carries no feature tag, because the rate looked too small to bother with.
- Retry caps on Luna paths are looser than on frontier paths, against a far larger call count.
- Nobody has ever priced the system prompt on your highest-volume surface.
Free, no card, no account
Run the free Cost Leak Scan
It shows your most expensive conversation before you install anything.
Mistakes that cost the most
| Mistake | Why it hurts | Do instead |
|---|---|---|
| Skipping attribution because the rate rounds to nothing. | Volume is the multiplier, and untagged spend is where an unexplained line item grows. | Tag Luna traffic exactly as you tag frontier traffic. The tag itself costs nothing. |
| Choosing between Luna and nano on price. | They share an input rate and differ by five cents on output, so cost barely separates them. | Test both on your own task, pick on output quality, then instrument whichever one ships. |
| Leaving a long system prompt on a cheap high-volume path. | At ten million calls, a thousand prompt tokens is $2,000 a month at this input rate. | Price the prompt directly: calls times tokens times $0.20 per million, then judge its length. |
Run this check tonight
- Find your monthly Luna call count. It usually sits far above your frontier count.
- Multiply prompt tokens by that call count by $0.20 per million. That's your fixed prompt cost.
- Check whether Luna traffic carries feature tags at all.
- Compare attempts to successful results on Luna paths specifically.
- Confirm the cached rate is in use. Luna carries $0.02 per million, a 90% discount.
Ten million calls on the cheapest tier
Illustrative example
A classification step running 10 million calls a month, each sending 700 input tokens and returning 50 output tokens.
$2,000 a month, and $800 of it disappears by shortening a prompt nobody has read. The tier choice already saved $23,000, and the prompt is the part still on the table.
Every number, with its confidence and source
| Figure | What it means | Confidence | Source |
|---|---|---|---|
| $0.20 / $1.20 per million | GPT-5.6 Luna input and output rates | calculated | Culpa price book row for openai/gpt-5.6-luna, $0.0002 and $0.0012 per 1k, read off the OpenAI pricing page on 2026-08-01. Pending seed into the ledger migrations. |
| $600 to $1,400 | modelled monthly input cost at ten million calls, trimmed prompt versus untrimmed | estimated | Both endpoints from the teardown arithmetic at the real Luna input rate. A range because call volume and prompt size are modelled. |
What a generic answer can’t know
A rate card can't tell you that your cheapest model carries your largest call count, or that most of its input is a system prompt written a year ago and never revisited. That needs per-call token data joined to your features. Culpa records it on your infrastructure, keeps your prompts and responses there, and counts the calls to run your plan.
Questions founders ask next
How much does GPT-5.6 Luna cost?
$0.20 per million input tokens and $1.20 per million output, with cached input at $0.02. That makes it the cheapest OpenAI tier, fractionally under GPT-5.4 nano on the output side.
Luna or GPT-5.4 nano?
They share the same $0.20 input rate and differ by five cents per million on output, so cost barely separates them. Choose on output quality for your specific task and instrument whichever one you ship.
Can a model this cheap really matter to my bill?
Yes, because it usually runs at the highest call volume in the product. At ten million calls a month a seven-hundred-token prompt costs $1,400, and nobody reviews it precisely because the per-call rate looks negligible.
Does the $0.20 rate hold on long prompts?
No. OpenAI publishes a long-context tier where input doubles and output rises by half, and the rate above is the short-context one. That tier reaches every full-size OpenAI model, though the cache-write price it also lists really is unique to the 5.6 family. Culpa's price book carries neither. Re-checked 2026-08-03.
On your infrastructure
Culpa runs on your infrastructure. Your prompts and responses never leave it. Culpa counts calls to run your plan, and it fails open, so if it ever breaks your app keeps running.
How Culpa works
Find the culprit. Not just the total.
Your dashboard shows what you spent. It stops short of who spent it. Culpa shows the conversation, the user and the feature behind it.
Your prompts stay local.
Culpa runs on your own infrastructure. What you send to a model reaches us at no point.
Every dollar has a name.
Follow any charge to the conversation, the user, the feature and the customer behind it.
See the bill before it lands.
Cost your next feature before you ship it. You get the likely bill and the worst case, at best, median, p90 and p99.
Three steps to your first answer.
Change one base URL.
Or drop in the Python or TypeScript library.
Find your most expensive conversation.
In the first session, not the first week.
Cost your next feature before you ship it.
Why the bill went up
Example dashboardCalls traced
418,209
across 3 projects
Spend this week
$378.41
+ $182 vs last week
Failed calls
312
74% retried, and you paid for all of them
+ $182 this week traced to one culprit
Spend over 14 days
Most expensive users
Next week forecast
Graded against reality. Accuracy shown as results land.
Keep reading
Sources: OpenAI API pricing. Last reviewed 2026-08-01, rates effective 2026-08-01. Plain text version.