Back to the index
30Small and fastProvisional score
GPT-5.6 Luna
GPT-5.6 Luna
Still cheap and still capable, but GPT-6 Luna costs half as much.
Updated
The read
Luna was the cheap end of the GPT-5.6 line, with a million-token window and the family's structured-output discipline at a fraction of its siblings' price. GPT-6 Luna now costs half as much again, and in October ChatGPT moved its free users to the newer model. The two score about the same, so the older one survives mainly where prompts were tuned against it.
Where it shines
- Million-token context at a low price
- Rated slightly higher than GPT-6 Luna by human voters
- Same tool-calling behaviour as its larger siblings
Worth knowing first
- GPT-6 Luna costs half as much
- Reasoning depth drops off sharply on multi-step problems
Index score
83.0
Provisional. This score is read from the product's public capability, not from the one-prompt rebuild the fully reviewed entries went through, so treat it as a placing rather than a verdict.
Craft80
Speed92
Control84
Value86
- Rank
- 30 of 73
- Builds
- LLMs
- Category
- Small and fast
- Output
- Text and code over an API
- Pricing
- Paid only
- Best for
- Existing high-volume jobs tuned against it
Scores are our own editorial judgement, weighted craft 55, speed 10, control 15 and value 20.
Compare it with
Also worth a look
Full index- 11Claude Haiku 5.5Claude Haiku 5.5ProvisionalA small model priced like a rounding error that scores where mid-tier models did last quarter.Small and fastPaid only87.9
- 22GPT-6 LunaGPT-6 LunaProvisionalThe cheap end of GPT-6, and the model free ChatGPT users now talk to.Small and fastPaid only84.9
- 49Mistral Small 4Mistral Small 4ProvisionalThe best small open model to build a product on, and the licence has no catch.Small and fastFree tier78.1
