GPT-6 Luna

GPT-6 Luna is OpenAI's small, inexpensive GPT-6 model, released alongside GPT-6 Sol on 22 September 2026. The API identifier is gpt-6-luna .

The price is the main attraction for my automated workflows: uncached input is half GPT-5.6 Luna's price, and output drops from $1.20 to $0.50 per million tokens. That makes it a candidate for the many small calls in AI Paper Podcasts, though lower token prices alone do not establish the cost or quality of a complete episode .

My GPT 6 Sol and Luna article covers the wider release, developer-reported benchmarks and a hands-on Sol example. This note keeps Luna's API details in one place.

Availability and reasoning

Specifications checked against OpenAI's documentation on 25 September 2026 :

It supports the Responses and Chat Completions APIs, streaming, structured outputs and function calling. Use Responses for tools with reasoning enabled: Chat Completions function calling requires reasoning_effort="none" .

Pricing and Cache Settings

OpenAI Standard API prices, verified 25 September 2026, in USD per million tokens. The first column applies to prompts of at most 272K input tokens; longer prompts use the second column for the entire request .

Cache reads cost 10% of the uncached input rate; cache writes cost 1.25 times that rate. Batch and Flex are half Standard prices, while Fast mode doubles the applicable rates. A longer reasoning trace can still increase the bill even when the model's token rates are low .

References

Changelog \textbar OpenAI API. https://developers.openai.com/api/docs/changelog.

OpenAI. Pricing. Living documentation. Accessed 8 October 2026. URL: https://developers.openai.com/api/docs/pricing (visited on 2026-10-08).

OpenAI. GPT-6 Luna. Living documentation. Accessed 8 October 2026. URL: https://developers.openai.com/api/docs/models/gpt-6-luna (visited on 2026-10-08).