GPT-6 Luna
Save 70%Officialopenai/gpt-6-luna
The fast, low-cost tier of GPT-6 for everyday work at scale.
- Context window
- 1.1M
- Max output
- 128K
- Input modalities
- text · image
- Output modalities
- text
- Capabilities
- Vision · Reasoning · Tool calling · Coding · Prompt cache · Paid cache writes · Normalize Openai Envelope · Strip Responses Summary Input
- Upstream release
- 2026-09-22
- RouteMux name
- openai/gpt-6-luna
Try GPT-6 Luna here
Tune the request, see the current price ceiling, and inspect the real response without leaving this model page.
Sign in first to upload images to this vision model.
Your result will appear here
Enter an input on the left. The same RouteMux Playground execution and billing path powers this embedded test.
What is GPT-6 Luna
GPT-6 Luna is the volume tier of OpenAI's GPT-6 family, released in September 2026 alongside GPT-6 Sol. It carries improvements from GPT-6 Astra into a model built for speed and throughput, and is meant for work that runs constantly: extraction, classification, routing, first-pass summarization, and the cheap legs of an agent loop.
It is not a stripped-down model. It accepts text and image input, supports tool calling and explicit prompt caching, and exposes reasoning effort as a request parameter from none all the way to max, so the same deployment can answer trivial turns with no reasoning and still think hard when a step needs it.
Through RouteMux it is $0.03 per 1M input tokens against OpenAI's $0.10, and $0.15 against $0.50 on output. It is reachable over OpenAI Chat Completions, OpenAI Responses, and Anthropic Messages, so switching an existing client to Luna is a model-string change.
GPT-6 Luna API pricing
Source: openai_x0.30 · Verified 2026-09-23
Fast $0.06 / $0.3 /M
Cache read $0.003/M · Cache write $0.0375/M
| Tier | Input | Cache read | Cache write | Output |
|---|---|---|---|---|
| Standard ≤272K | $0.03 | $0.003 | $0.0375 | $0.15 |
| Standard >272K | $0.06 | $0.006 | $0.075 | $0.225 |
| Priority ≤272K | $0.06 | $0.006 | $0.075 | $0.3 |
Estimate cost
Limited time20% extra Credits on your first top-up — up to 200 bonus Credits
Applies to your first top-up within 72 hours of signing up. Minimum $20, once per account.
Key capabilities
Built for volume
Priced and tuned for high-throughput work where per-million cost matters more than peak capability.
Reasoning from none to max
Effort is a request parameter; set none for plain completions or raise it for the occasional hard step.
Vision and tools included
Image input, tool calling and explicit prompt caching are all supported.
Three protocols
OpenAI Chat, OpenAI Responses, and Anthropic Messages all route to this model.
Interface and protocols
Use any of these IDs to call this model via the API.
openai/gpt-6-lunaReplace the ROUTEMUX_KEY placeholder with your API key. Create one →
from openai import OpenAI
client = OpenAI(
base_url="https://api.routemux.com/v1",
api_key="<ROUTEMUX_KEY>",
)
completion = client.chat.completions.create(
model="openai/gpt-6-luna",
messages=[{"role": "user", "content": "What is the meaning of life?"}],
)
print(completion.choices[0].message.content)Performance (last 30 days)
GPT-6 Luna FAQ
›What is GPT-6 Luna?
The fast, low-cost tier of GPT-6 for everyday work at scale.
›How large is the context window of GPT-6 Luna?
GPT-6 Luna supports a context window of up to 1.1M tokens, with up to 128K output tokens per request.
›How much does the GPT-6 Luna API cost?
Through RouteMux, GPT-6 Luna costs $0.03 for input and $0.15 for output (USD per 1M tokens), below the official list price.
›How do I call GPT-6 Luna via API?
GPT-6 Luna is OpenAI-compatible: point your base URL at RouteMux and set the model field to openai/gpt-6-luna — no code changes needed.
›Which input and output modalities does GPT-6 Luna support?
GPT-6 Luna accepts text, image as input and produces text as output.
›When was GPT-6 Luna released?
GPT-6 Luna was released on September 22, 2026.
›When should I use Luna instead of Astra?
Use Luna for high-volume, well-defined steps such as extraction, classification and routing. Astra is the flagship for tasks hard enough that a cheaper model would fail repeatedly.
›Can reasoning be turned off?
Yes. Luna accepts a reasoning effort of none, as well as low, medium, high, xhigh and max.
›Does the price change for very long inputs?
Yes. Requests whose input exceeds 272K tokens bill at the long-context rate. The pricing section above lists every tier.