See how ChatGPT, Perplexity and Google AI Overviews describe you today
Free AI Visibility AuditGPT-5.6 Luna is the fast, low-cost model in OpenAI's GPT-5.6 series, built for high-volume and latency-sensitive work such as chat, classification and lightweight agentic workflows.
OpenAIVerified 24/09/2026Released 09/07/202601 / decision
Each figure sits next to the middle of the field, so it can be read as dear or cheap, wide or narrow, rather than floating on its own. Compared against the 151 models we track, not against an absolute standard.
Capability
Not measured
No published benchmark scores yet
Input / 1M tokens
$0.20
field median $0.43
Output / 1M tokens
$1.20
field median $1.75
Context
1,050K
field median 524K
02 / overview
Four questions, answered separately, because somebody arrives at one of them rather than at the top of the page.
The job it was built for, and the job it is not for.
GPT-5.6 Luna is a general-purpose text and multimodal model positioned as the cheap, fast option within its series. It was built for work that runs at volume, classification, routing, chat responses and short agent steps, where latency and unit cost matter more than depth. It is not the model to reach for when a task needs sustained, careful reasoning, that is what its larger siblings in the series are for.
Where it sits in its line, and when a sibling is the better pick.
GPT-5.6 Luna was first seen in July 2026 as the fast tier of OpenAI's GPT-5.6 series. Pick it when a task runs thousands of times a day and each run is individually simple: tagging tickets, extracting fields, answering routine questions, driving a small agent loop. Pick a heavier model in the series when correctness on a single hard question matters more than throughput.
The API, the apps it powers, and what its limits let you do.
GPT-5.6 Luna is reached through the OpenAI API and sits behind OpenAI's own consumer and developer surfaces as a fast tier. It takes text, images and files as input, so a request can carry a scanned document or a screenshot alongside the prompt, and its context window is large enough to hold a full document set, a long support history or a sizeable codebase in a single call rather than chunking it. The output ceiling is generous enough for long structured returns, full reports or bulk JSON, not just short replies.
What changes because this exists, or why it does not.
The combination of a very large context window and a low per-token price changes what is worth reading. Work that previously meant retrieval, chunking and a fair amount of plumbing, feeding a model a long document set or a full conversation history, can now be handled by passing the whole thing in and paying very little for it. For teams running classification or agent loops at scale, the cost of being thorough has dropped to the point where it is no longer a design constraint.
Follows GPT-5.2 Pro. Superseded by GPT-5.6 Luna Pro. See the whole line.
03 / evidence
Split by category, so a strong number never hides a thin evidence base. Verified means we read it on the benchmark's own published results; a provider's claim about its own model is shown and labelled rather than dropped.
No published benchmark scores for this model yet.
We publish a score only where we can link the result to where it was published. Until a benchmark result for this model exists in a source we read, this section stays empty rather than being filled with a provider's marketing figure.
How we decide what counts as evidence04 / ledger
Every published row, grouped by category, each compared with the best published score on the same benchmark. 'Is 64% good' is a question nobody can answer; '26 points behind the leader' is one anybody can.
Nothing in the ledger yet.
Each row here carries a score, the benchmark it came from, what the leading model scored on the same test, and a link to the published result. Rows appear as results are published and read.
How we decide what counts as evidence05 / capability
Where this model is strong, and against how many peers. Ranks are against models with evidence in that category, not against everything we track: ranking against models nobody tested would rank who published, not who is better.
No category scores to shape yet.
A category score is the weighted mean of the benchmarks published for it. With no published rows there is nothing to average, and an empty chart drawn at zero would say something false.
How we decide what counts as evidence06 / cost
List API rates as last read from the provider, with the source on every row, plus every change we have recorded since we started tracking it.
| Charge | Price | Unit | Read on | Source |
|---|---|---|---|---|
| Input | $0.20 | per 1M tokens | 2026-09-24 | Check |
| Output | $1.20 | per 1M tokens | 2026-09-24 | Check |
GPT-5.6 Luna is priced at the cheap end of the current field, with output costing several times more than input, which is the usual shape for a fast tier. The practical effect is that reading is close to free and writing is where your bill accumulates, so it suits jobs that ingest a lot and return a little, and it is affordable enough to sit on every request rather than being reserved for the hard ones.
| Workload | Input / month | Output / month | Cost |
|---|---|---|---|
| A small product team | 20M tokens | 5M tokens | $10.00 |
| A busy support assistant | 200M tokens | 40M tokens | $88.00 |
| A document pipeline | 1000M tokens | 100M tokens | $320.00 |
List API rates, no caching and no batch discount, which both providers offer and which change the answer a great deal. Treat these as the ceiling, not the bill.
07 / specs
As listed by the provider's own catalogue and re-read every few hours. Anything absent is absent there too.
| Context window | 1,050,000 tokens |
|---|---|
| Maximum output | 128,000 tokens |
| Modalities | file, image, text |
| Released | 09/07/2026 |
| Status | Current |
| Catalogue identifier | openai/gpt-5.6-luna |
08 / lineage
What this model replaced, what replaced it, and what else its provider has in the field.
Also from OpenAI
09 / line
Every model in this family in release order, so a page from eight months ago says in one glance that two newer ones exist.
Came before
← GPT-5.2 ProCame after
GPT-5.6 Luna Pro →Ordered by release date and worked out from the naming, so a new member slots in as soon as its page exists. A retired model keeps its page and its place in the line.
10 / notes
What this model changes for a brand trying to be cited in AI answers, and every change we have logged since it launched.
What it changes for you
A cheap, long-context model with file and image input makes it economically sensible for an AI surface to read your whole documentation set, pricing page and comparison content on a single query rather than skimming a summary. That rewards brands whose material is complete, structured and machine-readable, and it punishes thin pages that only work when a model is forced to guess. Assume the model is reading everything you publish, not the first paragraph.
Buyers meet GPT-5.6 Luna through the OpenAI API, usually without knowing it, because it tends to be the model a product routes to for its everyday requests. It also sits underneath OpenAI's consumer chat surfaces as the fast responder, which is where most ordinary questions about vendors and products get answered.
Nothing published here yet. Changes appear within hours of a provider announcing them.
11 / questions
The things people ask about this model, answered from what is on this page rather than from anywhere else.
Models change what gets cited. We measure whether AI answers name your brand or your competitors across every assistant, and show you what to change.