OpenAI · in Borg

GPT-5.6 Luna

GPT-5.6 Luna is the cost-first, nano-like tier in the GPT-5.6 family.

Reasoning1,050,000 tokenstext, image

1,050,000 tokensContext window

3998 msMedian time to first token (gateway proxy)

17.2Output tokens per second (whole turn)

$0.2 / $1.2Vendor API price in / out · per 1M

Included in Borg

Download Borg Borg for Business

At a glance

Specification	Value
Provider	OpenAI
Family	GPT
Release date	2026-07-09 Vendor source
Context window	1,050,000 tokens Vendor source
Input modalities	text, image
Output modalities	text
Reasoning	Yes
Open weights	No
Knowledge cutoff	2026-02-16 Vendor source
Borg lane ids	`gpt-5.6-luna`, `openai/gpt-5.6-luna`
Versions in Borg	GPT-5.6 Luna · GPT-6 Luna
Median time to first token (gateway proxy) on Borg	3998 ms
Median tokens per second (end to end) on Borg	17.2 output tokens per second

Not published by the vendor: thinking modes, fusion roles.

What GPT-5.6 Luna is good for

GPT-5.6 Luna is the cost-first, nano-like tier in the GPT-5.6 family. It is cheaper than Terra and Sol and prioritizes repeated focused requests over hard reasoning. It accepts text and image inputs and generates text, with adjustable reasoning effort. The documented large context capacity is useful for substantial inputs, though unusually long API requests may be charged differently. Evaluate response quality against the neighboring tiers before committing a workflow, since matching context windows do not imply matching capability. Borg includes this lane in its subscription; Borg Fusion can route to it, or you can select its lane directly for a fixed model choice.

high-volume classification
quick text and image extraction
lightweight drafts
evaluating long-input workflows

How Borg uses it

GPT-5.6 Luna is available as a subscription catalog lane in Borg. The catalog does not publish which Fusion profiles route to this model, so no routing role is claimed here. You can pick it directly by setting the OpenAI-compatible gateway model field to a listed lane id. Check the current catalog for availability before building a workflow around a lane.

model: "gpt-5.6-luna"

Measured on Borg

Metric	Median	95th percentile	Samples
Time to first token (gateway proxy)	3998 ms	20334 ms	22631
Tokens per second (end to end)	17.2 output tokens per second	47.8 output tokens per second	605485

Time to first token (gateway proxy)

Lower is fasterFleet median 3971.5 ms

GPT-5.6 Luna3998 ms

Output tokens per second (whole turn)

Higher is fasterFleet median 31.9 tokens/s

GPT-5.6 Luna17.2 tokens/s

Time to first token (gateway proxy): time until the first streamed response event reaches the gateway; our closest available measure of time to first token
Tokens per second (end to end): tokens per second, end to end: provider-reported output tokens divided by the whole turn's elapsed time, including queueing and reasoning; not a streaming per-token rate

Completion speed covers the whole turn including queueing and reasoning time; it is not a streaming tokens-per-second figure.

Measured on Borg: SearchAtlas AI Gateway production request logs. 30-day window ending 2026-10-08; 752719 requests. Minimum 100 samples per published request metric.

Borg coding bench

Hard-graded with tests plus lint, no LLM judge; 3 tasks, 15 runs, 2026-09-20.

Metric	Pass rate
Overall pass rate	73%
Hidden-test pass rate	73%

Source: Borg coding bench, hard-graded by tests and ruff; no LLM judge

Pricing

Included in Borg subscription. Vendor API list price: $0.2 input / $1.2 output per 1M tokens. Vendor price source.

Standard global rates for prompts up to 272,000 input tokens. Above 272,000 input tokens, input and cached-input rates are 2× and output rates are 1.5× for the full request. Vendor pricing terms

Compare subscription and API pricing · Cost Savings Calculator

FAQ

What is GPT-5.6 Luna?

GPT-5.6 Luna is listed in Borg's subscription catalog as gpt-5.6-luna. GPT-5.6 Luna is the cost-first, nano-like tier in the GPT-5.6 family.

Is GPT-5.6 Luna in Borg?

Yes. GPT-5.6 Luna is listed in Borg's subscription catalog under the lane gpt-5.6-luna. You can choose the lane directly, or explore the available Fusion profiles. Availability and model details can change; check the current catalog before deployment.

What context window does GPT-5.6 Luna have?

The published context window is 1,050,000 tokens. Consult the cited sources for details and current limits. Vendor source

Is GPT-5.6 Luna good for coding?

No vendor-published coding evaluation is listed here. Its listed uses include high-volume classification, quick text and image extraction, lightweight drafts, evaluating long-input workflows. Try the lane for your own tasks, and compare the output with your requirements before choosing a default.

How fast is GPT-5.6 Luna on Borg?