Skip to content
VECTOR WIREAI INTELLIGENCE
UTC

GPT-5.6 Luna vs GPT-6 Luna

25 SHARED BENCHMARKS

Across 25 shared benchmarks, GPT-5.6 Luna scores higher on 14 and GPT-6 Luna on 11. The widest gap is AA-Omniscience · Non-hallucination, where GPT-6 Luna scores 23.3 against 7.4. GPT-6 Luna is the cheaper of the two on tracked API pricing ($0.10 against $0.20 per million input tokens).

OPENAIVSOPENAI25 SHARED14–11 HEAD-TO-HEADNEWEST SCORE ADDED

At a glance

MakerOpenAIOpenAI
Released
Price per 1M tokens input / output$0.20 / $1.20$0.10 / $0.50
Cost of 1M in + 1M out$1.40$0.60 2.3× less
Head-to-head of 25 shared benchmarks14 wins11 wins
Scores tracked independently verified183 43 ◆102 33 ◆

Release dates per Artificial Analysis. Prices: Artificial Analysis for GPT-5.6 Luna; Artificial Analysis for GPT-6 Luna. ◆ = independently verified score.

Where each leadsby capability area · benchmark wins

AreaGPT-5.6 LunaWINSGPT-6 Luna
Reasoning50GPT-5.6 Luna leads 5 of 5 · widest: LiveBench · Reasoning 85.6 vs 81.8
Coding13GPT-6 Luna leads 3 of 4 · widest: LMArena · WebDev 1520 vs 1593
Agentic11Even, 1–1 of 2
Factuality02GPT-6 Luna leads 2 of 2 · widest: AA-Omniscience · Non-hallucination 7.4 vs 23.3
Instruction Following10GPT-5.6 Luna leads 1 of 1 · widest: LiveBench · Instruction Following 60.1 vs 55.9
Long Context10GPT-5.6 Luna leads 1 of 1
Math01GPT-6 Luna leads 1 of 1
Multimodal10GPT-5.6 Luna leads 1 of 1 · widest: MMMU-Pro 78.6 vs 75.5

Biggest gaps

GPT-5.6 Luna pulls furthest ahead on

  1. GDPVal47.2 vs 43.4
  2. LiveBench · Instruction Following60.1 vs 55.9
  3. LiveBench · Coding82.9 vs 79

GPT-6 Luna pulls furthest ahead on

  1. AA-Omniscience · Non-hallucination23.3 vs 7.4
  2. LMArena · WebDev1593 vs 1520

Every shared benchmark25 · grouped by area

Reasoning 5

BenchmarkGPT-5.6 LunaMARGINGPT-6 Luna
ARC-AGI-30.2◆ 0.1
ARC-AGI-259.6 ◆◆ 59.3
CritPt20.619.4
LiveBench · Reasoning85.6 ◆◆ 81.8

Coding 4

BenchmarkGPT-5.6 LunaMARGINGPT-6 Luna
LMArena · WebDev1520 ◆◆ 1593
LiveBench · Coding82.9 ◆◆ 79
SciCode53.654.6

Agentic 2

BenchmarkGPT-5.6 LunaMARGINGPT-6 Luna
GDPVal47.243.4
Terminal-Bench 4.011.612.6

Instruction Following 1

Long Context 1

BenchmarkGPT-5.6 LunaMARGINGPT-6 Luna
AA-LCR83.783.3

Math 1

BenchmarkGPT-5.6 LunaMARGINGPT-6 Luna
LiveBench · Mathematics87.2 ◆◆ 89.1

Multimodal 1

BenchmarkGPT-5.6 LunaMARGINGPT-6 Luna
MMMU-Pro78.675.5

Other shared benchmarks 8

BenchmarkGPT-5.6 LunaMARGINGPT-6 Luna
ARC-AGI-190.7 ◆◆ 86.7
livebench_language72.6 ◆◆ 73.8
OSWorld 2.045.652.7

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.