GPT-6 Luna

by OpenAI - closed weights. The efficiency tier of GPT-6: “our most efficient model for focused, high-volume tasks”. Hub: gpt-5-and-gpt-6-family.

Item	Value (OpenAI API docs)
API model ID	gpt-6-luna
Released	2026-09-22 (with gpt-6-dot-sol)
Context / output	1,050,000 window, 922k max input, 128k max output
Knowledge cutoff	2026-05-18
Pricing	token-based; the cheapest GPT-6 tier
Reasoning effort	none, low, medium (default), high, xhigh, max
Modalities	text and image in, text out
Endpoints	Chat Completions, Responses, Batch

Use

Classification, summarization, extraction and other high-volume automation where cost and latency matter more than peak capability. Same 1M-token window as the larger tiers.

Notes

• GPT-5.6 also had a Luna tier (gpt-5.6-luna, 2026-07-09); OpenAI’s deprecation page names gpt-5.6-luna as the replacement for gpt-4.1-nano (shutdown 2026-10-23).
• ChatGPT availability of Luna: unverified.

Sources

• https://developers.openai.com/api/docs/models/gpt-6-luna (2026-09-30)
• https://developers.openai.com/api/docs/changelog (2026-09-30)
• https://developers.openai.com/api/docs/deprecations (2026-09-30)