Official commercial channel
GPT-5.6 Luna API
The fastest, cheapest tier of GPT-5.6 — with none of the window taken away.
Context
1.05M
Max output
128K
Serving precision
Vendor-managed
Licence
Proprietary
GPT-5.6 Luna is the fastest and cheapest of the three tiers, built for high-volume workloads at vendor list pricing of $1 per million input and $6 per million output. It shares the 1,050,000-token context window, 128,000-token output ceiling and knowledge cut-off with Sol and Terra, and accepts text and image input. In other words, the discount is paid for in reasoning depth, not in window — batch-processing long documents does not force a model change.
Proprietary, provided through the vendor’s official commercial licensing. Contract, invoice, audit trail — not an account pool and not a reverse-engineered client protocol.
Specifications
- Model ID
- gpt-5.6-luna
- Published by
- OpenAI
- How it is served
- Official commercial channel
- Licence
- Proprietary
- Context
- 1.05M
- Max output
- 128K
- Serving precision
- Vendor-managed
- Input modalities
- Text · Image
- Released
- 2026-07-09
- Price
- See pricing
What it is good at
- A fifth of Sol’s list price
- $1 input and $6 output per million tokens. On volume-driven pipelines that gap usually moves unit economics further than any amount of prompt tuning.
- The long window is intact
- 1.05M context and 128K output, identical to the top tier. That is unusual in a budget tier — most vendors get there by cutting the window.
- The default for high concurrency
- Explicitly built for high-volume workloads. Front-line classification, bulk cleaning and first-response support — high request count, low per-request difficulty — is exactly its shape.
Where it fits
- High-concurrency classification, extraction, cleaning and labelling
- The front end of chat and support products
- Summarisation of long documents at scale
- The cost floor in a multi-tier routing setup
Three steps
Change one line of base_url, put the model ID below in the `model` field, and leave the rest of your code alone.
model: "gpt-5.6-luna"
curl https://api.nexaix.net/v1/chat/completions \
-H "Authorization: Bearer $NEXAIX_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.6-luna",
"stream": true,
"messages": [
{ "role": "user", "content": "Summarise this contract." }
]
}'import os
from openai import OpenAI
client = OpenAI(
# the only new line
base_url="https://api.nexaix.net/v1",
api_key=os.environ["NEXAIX_API_KEY"],
)
# everything below is your existing code
stream = client.chat.completions.create(
model="gpt-5.6-luna",
stream=True,
messages=[
{"role": "user", "content": "Summarise this contract."}
],
)
for chunk in stream:
print(chunk.choices[0].delta.content or "", end="")import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.nexaix.net/v1", // the only new line
apiKey: process.env.NEXAIX_API_KEY,
});
// everything below is your existing code
const stream = await client.chat.completions.create({
model: "gpt-5.6-luna",
stream: true,
messages: [
{ role: "user", content: "Summarise this contract." }
],
});
for await (const chunk of stream) {
process.stdout.write(chunk.choices[0]?.delta?.content ?? "");
}About this model
Is Luna good enough?
It depends on how you tier the work. Most production requests are simple and deterministic; routing them to Luna can drop total cost by an order of magnitude, with measured failures escalated to Terra or Sol. Measure with your own eval before setting the threshold.
How much faster is it?
It is the fastest of the three. Actual latency depends on input length, output length and load at the time — test with your real request shapes rather than trusting a marketing figure.
Will I get rerouted at capacity?
No. At capacity you get a standard 429 with retry guidance and you decide whether to queue, back off or change tier. Silently swapping the model is something we refuse to do.
Also worth a look
- Official commercial channelClaude Opus 5A million tokens that never degrade, five-level effort control, zero data retention.claude-opus-5
- Official commercial channelClaude Fable 5Anthropic’s tier above Opus — the one you pick when the ceiling is the point.claude-fable-5
- Official commercial channelGPT-5.6 SolThe frontier tier of the GPT-5.6 family: 1.05M context, built for long-horizon agents.gpt-5.6-sol
Run your own eval on it
Take the trial credit, run the evals you already have against this model, and then decide whether it gets your production traffic.
