gpt-6-luna API — per-token pricing and code examples
Price per 1M tokens
Cache rates apply when the provider reports cache tokens.
80% offAbout this model
gpt-6-luna is served through OpenAI Chat Completions, OpenAI Responses, Anthropic Messages. You pay $0.01978 per 1M input tokens and $0.0989 per 1M output tokens. That is 80% below the official price of $0.10 and $0.50. Cache reads cost $0.001978 per 1M tokens when the provider reports cached tokens. It is included in budget packs from $1.99 up.
When GPT-6 Luna is the right pick
OpenAI describes GPT-6 Luna as its most efficient model for focused, high-volume tasks. Think of support triage, tagging thousands of records, short rewrites, or a router step that decides which tool an agent should call next. Those jobs run many times a day, and Luna is the member of the GPT-6 family that OpenAI aims at that kind of steady traffic.
The model page lists a context window of 1,050,000 tokens, of which up to 922,000 can be input, and up to 128,000 output tokens per response. GPT-6 Luna reads text and images and answers in text only. OpenAI released it on September 22, 2026 together with GPT-6 Sol, and a changelog entry from September 25 notes a fix to image encoding that had weakened image understanding in both models. If you tested Luna on screenshots in its first days and were disappointed, it is worth trying again.
Reasoning effort and tool calls on Luna
OpenAI's model page for GPT-6 Luna lists these effort levels: none, low, medium, high, xhigh and max, with medium as the default. Because none is on that list, you can run Luna without any reasoning step for the simplest calls. On /v1/chat/completions the field is reasoning_effort; on /v1/responses it is reasoning.effort.
There is one rule that catches people moving from older OpenAI models. According to the Luna model page, function calling through Chat Completions works only when reasoning_effort is none. OpenAI documents tool use for Luna on its Responses API without that restriction. On Cent AI, the /v1/responses format is available for Luna.
OpenAI's migration guide also says that whenever reasoning effort is not none, you should drop temperature, top_p and top_logprobs, plus logprobs on Chat Completions. Old sampling settings left in a config file are a common cause of failed requests after a model switch.
{
"model": "gpt-6-luna",
"reasoning_effort": "none",
"messages": [{ "role": "user", "content": "Label this ticket: billing, bug or other" }]
}
What it costs on Cent AI
Luna is billed per token from your prepaid USD balance, using the four rates in the price box on this page. When Luna reasons, OpenAI reports those hidden tokens in usage.output_tokens_details.reasoning_tokens, and Cent AI bills them at the output rate. So a ticket labelled at medium costs more than the same ticket at none, even when the visible answer is one word. For bulk classification, test the lowest effort that still gives correct labels before you scale up.
OpenAI lists prompt caching as supported for GPT-6 Luna. Cent AI applies the cache read and cache write rates only to the cached tokens OpenAI reports in the usage of each response, never to a guess. Keeping your fixed instructions at the start of every prompt gives the cache the best chance to match.
Using it from your tools
GPT-6 Luna is available on /v1/chat/completions, /v1/responses and /v1/messages. Cent AI does not change your parameters silently, so an option that Luna does not accept comes back as an upstream_error with the model's status code, for example 400; Cent AI does not drop it for you. OpenAI suggests leaving room of about 25,000 tokens for reasoning and output; when the limit is hit, a Responses call comes back with status: "incomplete" and the reason max_output_tokens. Watch for that status in high-volume jobs, because a cut-off label is easy to miss.
Specification
- Context window
- —
- Input and output
- Text
- Endpoints
- /v1/chat/completions, /v1/responses, /v1/messages
- Smallest pack
- $1.99
Latency and error rate per model are on the status page. status.cent-ai.dev
What a request costs
Open the full calculatorCost per request
Estimate
gpt-6-lunaAn estimate at price version 8. Your bill comes from the usage the provider reports for each request.
Save $0.0191Code
curl https://api.cent-ai.dev/v1/chat/completions \
-H "Authorization: Bearer $CENT_AI_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "gpt-6-luna", "messages": [{"role": "user", "content": "Hello"}]}'Set CENT_AI_KEY to a key from the dashboard. OpenAI SDKs use https://api.cent-ai.dev/v1; Anthropic SDKs use https://api.cent-ai.dev.
Price history
Version 8, in effect since Nov 19, 2026.
Effective
per 1M tokens
Inputold $0.024 new $0.01978Cache readold $0.0024 new $0.001978Cache writeold $0.03 new $0.024725Outputold $0.12 new $0.0989Effective
per 1M tokens
Inputold $0.01 new $0.024Cache readold $0.0022 new $0.0024Cache writeold $0.01 new $0.03Outputold $0.05 new $0.12Effective
per 1M tokens
Inputold $0.01 new $0.01Cache readold $0.001 new $0.0022Cache writeold $0.01 new $0.01Outputold $0.05 new $0.05Effective
per 1M tokens
Inputold $0.01 new $0.01Cache readold $0.001 new $0.001Cache writeold $0.0125 new $0.01Outputold $0.05 new $0.05Effective
per 1M tokens
Inputold $0.039 new $0.01Cache readold $0.0039 new $0.001Cache writeold $0.04875 new $0.0125Outputold $0.195 new $0.05Effective
per 1M tokens
Inputold $0.05 new $0.039Cache readold $0.005 new $0.0039Cache writeold $0.0625 new $0.04875Outputold $0.25 new $0.195Effective
per 1M tokens
Inputold $0.005 new $0.05Cache readold $0.0005 new $0.005Cache writeold $0.00625 new $0.0625Outputold $0.025 new $0.25
Compare with
- deepseek-v4.1-flashInput$0.015
- qwen3.8-flashInput$0.015
- deepseek-v4-flashInput$0.015
- glm-5.3-flashInput$0.015
Questions
How much does gpt-6-luna cost on Cent AI?
$0.01978 per 1M input tokens and $0.0989 per 1M output tokens. Cache reads cost $0.001978 and cache writes $0.024725 per 1M tokens. Each request is rounded up once.
Is there a minimum charge?
A successful request costs at least $0.00002. A request that ends early, because the client left or the output was cut off, pays only for its tokens.
How do I pay?
Top up a prepaid USD balance with QRIS or USDT/USDC, or buy a budget pack once. Nothing renews automatically.
What if the provider does not report usage?
The request is billed from an estimate and marked estimated on the Requests page. An estimate never contains cache tokens.
Which endpoint do I call?
/v1/chat/completions, /v1/responses, /v1/messages. The OpenAI base URL is https://api.cent-ai.dev/v1.
Try gpt-6-luna with one key
One prepaid balance for every model. Pay with QRIS or USDT/USDC.