Enter opens the highlighted result. Esc closes.

GPT-5.6 Luna vs GPT-6 Luna

GPT-5.6 Luna leads GPT-6 Luna on coding, 58.0 to 57.3 on the Vibe Coding Index, and GPT-6 Luna costs 2.4x less per output token ($0.50 vs $1.20 per million). Updated October 22, 2026; 96 benchmarks have a result for both.

Replace GPT-5.6 Luna Replace GPT-6 Luna
GPT-5.6 LunaOpenAI
Coding
58.0
Overall
61.6
Output $/M
$1.20
57
GPT-6 LunaOpenAI
Coding
57.3
Overall
59.7
Output $/M
$0.50
38
vs boards won of 96 shared, 1 tied

GPT-5.6 Luna or GPT-6 Luna: which to pick

GPT-5.6 Luna and GPT-6 Luna are level for coding, 58.0 against 57.3 on the Vibe Coding Index, so price and context decide.

Pick GPT-6 Luna for output tokens 2.4x cheaper ($0.50 vs $1.20 per million) and work measured by LiveBench coding or LMArena WebDev, where it leads. Pick GPT-5.6 Luna for work measured by Code migration (Vals AI) or IOI (Vals AI), where it leads.

Model AGPT-5.6 LunaOpenAI58.0/1006 independent benchmarks$0.04 per solved taskModel BGPT-6 LunaOpenAI57.3/1005 independent benchmarksNo cost per solved task
Shared results
96
GPT-5.6 Luna only
24
GPT-6 Luna only
8
Like-for-like categories
8 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
GPT-5.6 Luna 58.0 from 6 benchmarks, GPT-6 Luna 57.3 from 5 benchmarks.
Widest category
Knowledge, 9.0 points apart on average over 7 shared benchmarks.
Widest single gap
FACTS Benchmark Suite: GPT-5.6 Luna 35.0% against GPT-6 Luna 63.3%, 28.2 points apart on a 0 to 100 scale.
Output price
GPT-5.6 Luna $1.20 against GPT-6 Luna $0.50 per million output tokens, 2.4x apart.
Context window
Both take 1.05M tokens.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryGPT-5.6 LunaBarsGPT-6 LunaTies
Coding611
Intelligence02
Agentic work471 tie
Reasoning121
Math42
Knowledge43
Web design09
General273
All shared boards57381 tie
CodingLike-for-likeSix of these boards feed the Vibe Coding Index611
Coding benchmarks for GPT-5.6 Luna and GPT-6 Luna
BenchmarkGPT-5.6 LunaGPT-6 LunaLeader
LiveBench coding73.6%84.5%GPT-6 Luna
LMArena WebDev15181578GPT-6 Luna
AA Coding Index71.4Not measuredOnly GPT-5.6 Luna measured
SWE-bench Verified (Vals AI)93.0%Not measuredOnly GPT-5.6 Luna measured
Terminal-Bench 4 (Vals AI)11.6%13.6%GPT-6 Luna
Vibe Code Bench (Vals AI)77.1%81.7%GPT-6 Luna
Code migration (Vals AI)44.5%42.5%GPT-5.6 Luna
ProgramBench (Vals AI)0.0%0.5%GPT-6 Luna
IOI (Vals AI)61.8%55.6%GPT-5.6 Luna
Terminal-Bench 2.1 (Vals AI)79.0%73.0%GPT-5.6 Luna
SWE-bench Multimodal (Vals AI)Not measured37.5%Only GPT-6 Luna measured
Vibe Code Bench 1-100 (Vals AI)22.6%Not measuredOnly GPT-5.6 Luna measured
SkillsBench (Vals AI)60.5%Not measuredOnly GPT-5.6 Luna measured
CyberBench (Vals AI)73.6%76.3%GPT-6 Luna
SciCode53.6%54.6%GPT-6 Luna
FrontierCode39.8%42.4%GPT-6 Luna
WeirdML60.9%Not measuredOnly GPT-5.6 Luna measured
ALE-Bench1667.41576.9GPT-5.6 Luna
LMArena Text: coding14661437GPT-5.6 Luna
LMArena Text: software and IT14631423GPT-5.6 Luna
LMArena WebDev: React15081573GPT-6 Luna
LMArena WebDev: HTML15201595GPT-6 Luna
LMArena WebDev: from an image15171571GPT-6 Luna
SciCode6.2%Not measuredOnly GPT-5.6 Luna measured
Neither model has a result on the other 19 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for GPT-5.6 Luna and GPT-6 Luna
BenchmarkGPT-5.6 LunaGPT-6 LunaLeader
AA Intelligence Index37.338.1GPT-6 Luna
Epoch Capabilities Index156.5156.8GPT-6 Luna
Agentic workLike-for-likePublic boards, shown as published, kept out of the index47
Agentic work benchmarks for GPT-5.6 Luna and GPT-6 Luna
BenchmarkGPT-5.6 LunaGPT-6 LunaLeader
AA Agentic Index42.1Not measuredOnly GPT-5.6 Luna measured
Design Arena: agents building web appsNot measured1191Only GPT-6 Luna measured
Design Arena: agents building full-stack appsNot measured1158Only GPT-6 Luna measured
Design Arena: agents building mobile appsNot measured1131Only GPT-6 Luna measured
LiveBench research51.0%43.5%GPT-5.6 Luna
LiveBench browser68.0%72.0%GPT-6 Luna
LiveBench long agentic30.4%48.1%GPT-6 Luna
SREBench (Vals AI)1.5%2.7%GPT-6 Luna
Finance Agent (Vals AI)55.0%49.9%GPT-5.6 Luna
Tax Agent Bench (Vals AI)24.8%19.9%GPT-5.6 Luna
Legal Research Bench (Vals AI)36.5%30.3%GPT-5.6 Luna
Harvey's Legal Agent Benchmark (Vals AI)Not measured4.2%Only GPT-6 Luna measured
Excel Modeling Benchmark (Vals AI)67.1%68.5%GPT-6 Luna
BioMysteryBench (Vals AI)61.5%61.5%Tie
Terminal-Bench-Science (Vals AI)0.0%4.3%GPT-6 Luna
Vals Science Index25.3%28.4%GPT-6 Luna
APEX-Agents43.0%44.3%GPT-6 Luna
BALROG45.6%Not measuredOnly GPT-5.6 Luna measured
SpireClimbNot measured0.0%Only GPT-6 Luna measured
Vending-Bench 2$4,095Not measuredOnly GPT-5.6 Luna measured
Neither model has a result on the other 22 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index121
Reasoning benchmarks for GPT-5.6 Luna and GPT-6 Luna
BenchmarkGPT-5.6 LunaGPT-6 LunaLeader
GPQA Diamond91.7%Not measuredOnly GPT-5.6 Luna measured
LiveBench reasoning85.6%81.8%GPT-5.6 Luna
MysteryMechanism (Vals AI)14.4%19.4%GPT-6 Luna
MMMU Pro (Vals AI)85.0%Not measuredOnly GPT-5.6 Luna measured
GPQA Diamond (Artificial Analysis run)91.1%Not measuredOnly GPT-5.6 Luna measured
Humanity's Last Exam (Artificial Analysis run)39.5%38.5%GPT-5.6 Luna
GPQA Diamond (Epoch AI run)91.6%90.5%GPT-5.6 Luna
ARC-AGI88.0%86.7%GPT-5.6 Luna
ARC-AGI-259.5%59.3%GPT-5.6 Luna
SimpleBench46.8%Not measuredOnly GPT-5.6 Luna measured
Chess puzzles43.0%36.0%GPT-5.6 Luna
Mystery game puzzles21.0%7.0%GPT-5.6 Luna
LMArena Text: hard prompts14511410GPT-5.6 Luna
LMArena Text: expert14781442GPT-5.6 Luna
LMArena Text: hard prompts in English14601410GPT-5.6 Luna
LMArena Vision: diagrams12731245GPT-5.6 Luna
LMArena Vision: homework12821261GPT-5.6 Luna
Kagi LLM Benchmark49.1%Not measuredOnly GPT-5.6 Luna measured
Neither model has a result on the other 5 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index42
Math benchmarks for GPT-5.6 Luna and GPT-6 Luna
BenchmarkGPT-5.6 LunaGPT-6 LunaLeader
OTIS Mock AIME98.3%98.9%GPT-6 Luna
FrontierMath, tiers 1 to 382.1%78.9%GPT-5.6 Luna
FrontierMath, tier 461.0%56.1%GPT-5.6 Luna
ProofBench60.0%64.0%GPT-6 Luna
LMArena Text: math14591409GPT-5.6 Luna
LMArena Text: mathematical work14701427GPT-5.6 Luna
Neither model has a result on the other 8 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index43
Knowledge benchmarks for GPT-5.6 Luna and GPT-6 Luna
BenchmarkGPT-5.6 LunaGPT-6 LunaLeader
MMLU Pro86.0%Not measuredOnly GPT-5.6 Luna measured
LegalBench (Vals AI)84.0%Not measuredOnly GPT-5.6 Luna measured
CorpFin (Vals AI)64.2%Not measuredOnly GPT-5.6 Luna measured
TaxEval (Vals AI)76.2%Not measuredOnly GPT-5.6 Luna measured
MortgageTax (Vals AI)67.3%Not measuredOnly GPT-5.6 Luna measured
MedCode (Vals AI)42.4%44.7%GPT-6 Luna
SimpleQA Verified41.0%41.4%GPT-6 Luna
LMArena Text: science14331386GPT-5.6 Luna
LMArena Text: medicine and healthcare13991352GPT-5.6 Luna
LMArena Text: legal and government14461400GPT-5.6 Luna
LMArena Text: business and finance14391388GPT-5.6 Luna
FACTS Benchmark Suite35.0%63.3%GPT-6 Luna
PRBench LegalNot measured42.9%Only GPT-6 Luna measured
Neither model has a result on the other 7 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index09
Web design benchmarks for GPT-5.6 Luna and GPT-6 Luna
BenchmarkGPT-5.6 LunaGPT-6 LunaLeader
Design Arena: websites12551271GPT-6 Luna
Design Arena: UI components12481295GPT-6 Luna
Design Arena: game dev11951258GPT-6 Luna
Design Arena: data viz12121217GPT-6 Luna
Design Arena: 3D11211218GPT-6 Luna
Design Arena: all code categories12301270GPT-6 Luna
Design Arena: SVG11931253GPT-6 Luna
Design Arena: ASCII art11421153GPT-6 Luna
LiveBench design76.0%82.5%GPT-6 Luna
GeneralLike-for-likePublic boards, shown as published, kept out of the index273
General benchmarks for GPT-5.6 Luna and GPT-6 Luna
BenchmarkGPT-5.6 LunaGPT-6 LunaLeader
Vals IndexNot measured52.3%Only GPT-6 Luna measured
LiveBench data analysis78.0%73.4%GPT-5.6 Luna
Public Benefits Bench (Vals AI)61.2%57.6%GPT-5.6 Luna
SAGE (Vals AI)44.2%48.1%GPT-6 Luna
MedScribe (Vals AI)84.4%83.7%GPT-5.6 Luna
Vals Multimodal Index69.1%Not measuredOnly GPT-5.6 Luna measured
Long context reasoning (AA-LCR)83.7%83.3%GPT-5.6 Luna
DTBench89.1%90.1%GPT-6 Luna
LMCA48.5%44.5%GPT-5.6 Luna
Surface Evolver Bench61.9%Not measuredOnly GPT-5.6 Luna measured
Furniture assembly42.5%44.2%GPT-6 Luna
LMArena Text14311389GPT-5.6 Luna
LMArena Text: instruction following14371408GPT-5.6 Luna
LMArena Text: creative writing13961363GPT-5.6 Luna
LMArena Vision12571216GPT-5.6 Luna
LMArena Text: English14411393GPT-5.6 Luna
LMArena Text: non-English14171381GPT-5.6 Luna
LMArena Text: Chinese14701427GPT-5.6 Luna
LMArena Text: German14551376GPT-5.6 Luna
LMArena Text: Japanese14151369GPT-5.6 Luna
LMArena Text: Korean14151359GPT-5.6 Luna
LMArena Text: French14581408GPT-5.6 Luna
LMArena Text: Spanish14461391GPT-5.6 Luna
LMArena Text: Russian14271392GPT-5.6 Luna
LMArena Text: Polish14051383GPT-5.6 Luna
LMArena Text: multi-turn14341392GPT-5.6 Luna
LMArena Text: longer queries14361408GPT-5.6 Luna
LMArena Text: writing and language14191384GPT-5.6 Luna
LMArena Text: entertainment and media14021364GPT-5.6 Luna
LMArena Vision: English12631204GPT-5.6 Luna
LMArena Vision: Chinese12941266GPT-5.6 Luna
LMArena Vision: reading text12641225GPT-5.6 Luna
LMArena Vision: humor1190Not measuredOnly GPT-5.6 Luna measured
LMArena Vision: creative writing12571205GPT-5.6 Luna
LMArena Document1457Not measuredOnly GPT-5.6 Luna measured
EQ-Bench 41156Not measuredOnly GPT-5.6 Luna measured
Neither model has a result on the other 12 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-5.6 Luna and GPT-6 Luna
BenchmarkGPT-5.6 LunaGPT-6 LunaLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

GPT-5.6 Luna$0.0008Fits in one request
GPT-6 Luna$0.0003Fits in one request

GPT-6 Luna costs less

Repository review

50K input + 3K output tokens

GPT-5.6 Luna$0.014Fits in one request
GPT-6 Luna$0.0065Fits in one request

GPT-6 Luna costs less

Agent loop

200K input + 10K output tokens

GPT-5.6 Luna$0.052Fits in one request
GPT-6 Luna$0.025Fits in one request

GPT-6 Luna costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of GPT-5.6 Luna and GPT-6 Luna
SpecificationOpenAI GPT-5.6 LunaOpenAI GPT-6 Luna
MakerOpenAIOpenAI
WeightsNo public weights listedNo public weights listed
ReleasedJuly 9, 2026September 22, 2026
Context window1.05M1.05M
Input $/Mtok$0.20$0.10
Output $/Mtok$1.20$0.50
Coding index58.057.3
Independent measurements behind it65
Cost per solved task$0.04Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-5.6 Luna better than GPT-6 Luna for coding?

GPT-5.6 Luna and GPT-6 Luna are level for coding, 58.0 against 57.3 on the Vibe Coding Index, so price and context decide. Pick GPT-6 Luna for output tokens 2.4x cheaper ($0.50 vs $1.20 per million) and work measured by LiveBench coding or LMArena WebDev, where it leads. Pick GPT-5.6 Luna for work measured by Code migration (Vals AI) or IOI (Vals AI), where it leads.

Which is cheaper, GPT-5.6 Luna or GPT-6 Luna?

GPT-5.6 Luna costs $0.20 per million input tokens and $1.20 per million output tokens; GPT-6 Luna costs $0.10 and $0.50. Cost per solved task is known for GPT-5.6 Luna only.

Which has the bigger context window, GPT-5.6 Luna or GPT-6 Luna?

Neither: both take 1.05M tokens.

How many benchmarks is the GPT-5.6 Luna vs GPT-6 Luna comparison based on?

96 benchmarks have a result for both models as of October 22, 2026. GPT-5.6 Luna's coding index rests on 6 independent benchmarks and GPT-6 Luna's on 5. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/gpt-5-6-luna-vs-gpt-6-luna" title="GPT-5.6 Luna vs GPT-6 Luna, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/gpt-5-6-luna-vs-gpt-6-luna">LLMBenchmark.ai, GPT-5.6 Luna vs GPT-6 Luna</a></p>