GPT-6 Luna
GPT-6 Luna is OpenAI's small, cheap GPT-6 model, released alongside GPT-6 Sol on 22 September 2026 (openaiAPIChangelog).
Input is half GPT-5.6 Luna's price, and output drops from $1.20 to $0.50 per million tokens (OpenAI). It's my go-to when I need a cost-efficient model.
The wider release, benchmarks and a hands-on example are in GPT 6 Sol and Luna. For a comparison with Anthropic's equivalent, see Claude Haiku 5.5.
Specs
| Specification | GPT-6 Luna |
|---|---|
| API identifier | gpt-6-luna |
| Input / output | Text and images / text |
| Context window | 1,050,000 tokens (922,000 max input) |
| Maximum output | 128K tokens |
| Knowledge cutoff | 18 May 2026 |
| Effort levels | none to max, default medium |
As of 8 October 2026 (OpenAI). For tool calls with reasoning on, use the Responses API: Chat Completions function calling requires reasoning_effort="none".
Pricing
USD per million tokens, verified 8 October 2026. Over 272K input tokens, the whole request pays the higher rate (OpenAI).
| Usage | Up to 272K | Over 272K |
|---|---|---|
| Input, uncached | $0.10 | $0.20 |
| Input, cache read | $0.01 | $0.02 |
| Cache write | $0.125 | $0.25 |
| Output, including reasoning | $0.50 | $0.75 |
Batch and Flex are half price. Fast mode is double.
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). ↩ 1 2
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). ↩