0

GPT-5.6 Luna

Restricted

The fastest, most cost-efficient tier of OpenAI's GPT-5.6 family, available in limited preview through the OpenAI API and Codex for approved organizations.

Publisher
OpenAI
Family
GPT
Context
1.1M
Input
Output
Providers
openai
Released
26 Jun 2026

Cite

Notes

Only stored in your browser.

Attribution

Attribution policy →
Context
1.1M
Weights
Closed
API only
Input
$0.20
/1M tok
Output
$1.20
/1M tok
Blended
$0.45
#312/ 494/1M · 3:1
Speed
0
#103/ 493tok/s
Cache read
$0.02
0.10x input
Cache write
$0.25
1.25x input

Price & speed via Artificial Analysis · ranked across tracked models · caching via OpenRouter

Reasoning effortIndexSpeed
max37.30 t/s
xhigh34.60 t/s
high32.10 t/s
medium25.00 t/s
low21.00 t/s
Non-reasoning15.50 t/s

Intelligence Index per reasoning-effort setting, via Artificial Analysis · pricing identical across tiers

Reported on 21 evals across 11 domains - best 91.1% on GPQA Diamond · #33 of 511

Reported eval scores

21
TypeSafe Workflow Evals: Agent trace observability76.1
TypeSafe Workflow Evals: Customer service71.4
TypeSafe Workflow Evals: Invoice processing67.8
TypeSafe Workflow Evals: Security incidents52.1
TaxEval v276.2
Finance Agent (v2)55
MortgageTax67.3
CorpFin v264.2
GPQA Diamond91.1
SciCode39.9
Humanity's Last Exam (HLE)39.5
Vals Index51.7
EQ-Bench Creative Writing v31828.7
EQ-Bench Longform Creative Writing75.5
Vals Multimodal Index69.1
ProofBench60
Vibe Code Bench v1.177.1
MedScribe84.4
MedCode42.4
SAGE44.2