GPT-5.6 Luna

GPT-5.6 Luna is the OpenAI model for jobs that run constantly and must answer fast: support chat, routing messages, pulling fields out of documents. OpenAI released it in July 2026 as the fastest and cheapest model of GPT-5.6, with enough reasoning for most routine requests. A few weeks after launch OpenAI cut its list price.

Your existing OpenAI SDK code works unchanged once the base URL points at Kadraris.

What you can use it for

• Labeling and sorting large batches of text
• First-response answers in a support widget
• Reading invoices, forms and emails into structured fields
• Short summaries of long threads or log files

Settings and limits

• Context window: 1,050,000 tokens; output up to 128K tokens.
• Tool calls: pass functions in the OpenAI tools format and run the returned call in your code. See the chat endpoint docs.
• For extraction, check the returned fields in code; any model can occasionally miss one.
• Keep prompts short and fixed so batch costs stay predictable.

Example request

from openai import OpenAI

client = OpenAI(
    base_url="https://kadraris.com/v1",
    api_key="sk-kd-...",
)
resp = client.chat.completions.create(
    model="gpt-5.6-luna",
    messages=[{"role": "user",
               "content": "Get invoice number, date and "
                          "total as JSON from this text."}],
)
print(resp.choices[0].message.content)

Change the field list to match your documents and validate the JSON before saving it.

Questions

What does GPT-5.6 Luna cost per token?

$0.216 / $1.30 per 1M tokens. With short prompts, a large batch costs very little, and you only pay for tokens you send and receive. Failed requests cost $0.

Can it handle coding work?

Small edits, snippets and explanations are fine. For multi-file changes or debugging, use GPT-5.6 Terra or a Sol model, which OpenAI aims at heavier coding.

Is there a newer Luna?

Yes, GPT-6 Luna is the next generation of the same tier. Test both on your prompts before switching.

GPT-5.6 Luna in your code

from openai import OpenAI

client = OpenAI(base_url="https://kadraris.com/v1", api_key="sk-kd-...")
resp = client.chat.completions.create(
    model="gpt-5.6-luna",
    messages=[{"role": "user", "content": "Suggest three opening shots for a short film about a cafe."}],
)
print(resp.choices[0].message.content)

curl https://kadraris.com/v1/chat/completions \
  -H "Authorization: Bearer $KADRARIS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "gpt-5.6-luna", "messages": [{"role": "user", "content": "Suggest three opening shots for a short film about a cafe."}]}'

import OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://kadraris.com/v1", apiKey: process.env.KADRARIS_API_KEY });
const resp = await client.chat.completions.create({
  model: "gpt-5.6-luna",
  messages: [{ role: "user", content: "Suggest three opening shots for a short film about a cafe." }],
});
console.log(resp.choices[0].message.content);