GPT-5 vs GPT-5.6 Luna
34 SHARED BENCHMARKSAcross 34 shared benchmarks, GPT-5 scores higher on 8 and GPT-5.6 Luna on 26. The widest gap is ARC-AGI-2, where GPT-5.6 Luna scores 59.6 against 9.9. GPT-5.6 Luna is the cheaper of the two on tracked API pricing ($0.20 against $1.25 per million input tokens).
At a glance
Release dates per Artificial Analysis. Prices: Artificial Analysis for GPT-5; Artificial Analysis for GPT-5.6 Luna. ◆ = independently verified score.
Where each leadsby capability area · benchmark wins
Biggest gaps
GPT-5 pulls furthest ahead on
- AA-Omniscience · Non-hallucination17.8 vs 7.4
- SimpleQA Verified50.1 vs 41
- SimpleBench56.7 vs 46.8
GPT-5.6 Luna pulls furthest ahead on
- ARC-AGI-259.6 vs 9.9
- CritPt20.6 vs 5.7
- FrontierMath Tier 458.5 vs 21.9
Every shared benchmark34 · grouped by area
Reasoning 5
Coding 5
Agentic 3
Factuality 3
Long Context 1
Math 2
Multimodal 3
Other shared benchmarks 12
Each score is the one the model's own page shows — the most authoritative tracked result for that benchmark, on the benchmark's standard methodology (source-attributed). ↓ marks lower-is-better metrics; ◆ an independently verified score. Open either model for its full surface, provenance and pricing. Quoted rates are the price-setter row we currently track for each model — its direct or vendor-official listing where one exists (direct, direct), otherwise the lowest tracked offer.