Enter opens the highlighted result. Esc closes.

Claude Opus 5 vs GPT-5.6 Luna

Claude Opus 5 leads GPT-5.6 Luna on coding, 83.0 to 58.0 on the Vibe Coding Index, and GPT-5.6 Luna costs 20.8x less per output token ($1.20 vs $25.00 per million). Updated October 22, 2026; 116 benchmarks have a result for both.

Replace Claude Opus 5 Replace GPT-5.6 Luna
Claude Opus 5Anthropic
Coding
83.0
Overall
81.6
Output $/M
$25.00
110
GPT-5.6 LunaOpenAI
Coding
58.0
Overall
61.6
Output $/M
$1.20
5
vs boards won of 116 shared, 1 tied

Claude Opus 5 or GPT-5.6 Luna: which to pick

Pick Claude Opus 5 for coding: it scores 83.0 on the Vibe Coding Index against 58.0 for GPT-5.6 Luna.

Pick GPT-5.6 Luna for output tokens 20.8x cheaper ($1.20 vs $25.00 per million) and a lower cost per solved task ($0.04 against $1.33).

Model AClaude Opus 5Anthropic83.0/1006 independent benchmarks$1.33 per solved taskModel BGPT-5.6 LunaOpenAI58.0/1006 independent benchmarks$0.04 per solved task
Shared results
116
Claude Opus 5 only
15
GPT-5.6 Luna only
3
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
Claude Opus 5 83.0 from 6 benchmarks, GPT-5.6 Luna 58.0 from 6 benchmarks.
Widest category
Agentic work, 20.8 points apart on average over 15 shared benchmarks.
Widest single gap
Terminal-Bench 4 (Vals AI): Claude Opus 5 53.5% against GPT-5.6 Luna 11.6%, 55.9 points apart on a 0 to 100 scale.
Output price
Claude Opus 5 $25.00 against GPT-5.6 Luna $1.20 per million output tokens, 20.8x apart.
Context window
Claude Opus 5 takes 1M tokens, GPT-5.6 Luna 1.05M.

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
CategoryClaude Opus 5BarsGPT-5.6 LunaTies
Coding212
Intelligence20
Agentic work150
Reasoning1601 tie
Math60
Knowledge101
Web design90
General312
All shared boards11051 tie
CodingLike-for-likeSix of these boards feed the Vibe Coding Index212
Coding benchmarks for Claude Opus 5 and GPT-5.6 Luna
BenchmarkClaude Opus 5GPT-5.6 LunaLeader
LiveBench coding90.8%73.6%Claude Opus 5
LMArena WebDev16891518Claude Opus 5
AA Coding Index78.071.4Claude Opus 5
LiveCodeBench (Vals AI)89.0%Not measuredOnly Claude Opus 5 measured
SWE-bench Verified (Vals AI)97.0%93.0%Claude Opus 5
Terminal-Bench 4 (Vals AI)53.5%11.6%Claude Opus 5
Vibe Code Bench (Vals AI)88.4%77.1%Claude Opus 5
Code migration (Vals AI)57.5%44.5%Claude Opus 5
ProgramBench (Vals AI)3.0%0.0%Claude Opus 5
IOI (Vals AI)84.3%61.8%Claude Opus 5
Terminal-Bench 2.1 (Vals AI)84.6%79.0%Claude Opus 5
Vibe Code Bench 1-100 (Vals AI)28.5%22.6%Claude Opus 5
SkillsBench (Vals AI)60.4%60.5%GPT-5.6 Luna
CyberBench (Vals AI)65.4%73.6%GPT-5.6 Luna
SciCode56.4%53.6%Claude Opus 5
FrontierCode53.4%39.8%Claude Opus 5
FrontierSWE52.0%Not measuredOnly Claude Opus 5 measured
WeirdML91.8%60.9%Claude Opus 5
ALE-Bench2164.61667.4Claude Opus 5
LMArena Text: coding15331466Claude Opus 5
LMArena Text: software and IT15271463Claude Opus 5
LMArena WebDev: React17111508Claude Opus 5
LMArena WebDev: HTML16801520Claude Opus 5
LMArena WebDev: from an image16621517Claude Opus 5
SciCode13.8%6.2%Claude Opus 5
Neither model has a result on the other 14 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for Claude Opus 5 and GPT-5.6 Luna
BenchmarkClaude Opus 5GPT-5.6 LunaLeader
AA Intelligence Index50.837.3Claude Opus 5
Epoch Capabilities Index163.9156.5Claude Opus 5
Agentic workLike-for-likePublic boards, shown as published, kept out of the index150
Agentic work benchmarks for Claude Opus 5 and GPT-5.6 Luna
BenchmarkClaude Opus 5GPT-5.6 LunaLeader
AA Agentic Index56.542.1Claude Opus 5
Design Arena: agents building web apps1266Not measuredOnly Claude Opus 5 measured
Design Arena: agents building full-stack apps1306Not measuredOnly Claude Opus 5 measured
Design Arena: agents building mobile apps1348Not measuredOnly Claude Opus 5 measured
Design Arena: agents building Android apps1257Not measuredOnly Claude Opus 5 measured
Design Arena: agents building Godot games1279Not measuredOnly Claude Opus 5 measured
Design Arena: agents building games1261Not measuredOnly Claude Opus 5 measured
Design Arena: PowerPoint slides1240Not measuredOnly Claude Opus 5 measured
LiveBench research88.9%51.0%Claude Opus 5
LiveBench browser84.0%68.0%Claude Opus 5
LiveBench long agentic62.1%30.4%Claude Opus 5
SREBench (Vals AI)12.2%1.5%Claude Opus 5
Finance Agent (Vals AI)58.6%55.0%Claude Opus 5
Tax Agent Bench (Vals AI)45.5%24.8%Claude Opus 5
Legal Research Bench (Vals AI)55.3%36.5%Claude Opus 5
Excel Modeling Benchmark (Vals AI)73.6%67.1%Claude Opus 5
BioMysteryBench (Vals AI)79.3%61.5%Claude Opus 5
Terminal-Bench-Science (Vals AI)27.1%0.0%Claude Opus 5
Vals Science Index47.9%25.3%Claude Opus 5
Time Horizon Index (Vals AI)18.8%Not measuredOnly Claude Opus 5 measured
CUA-bench (Vals AI)9.0%Not measuredOnly Claude Opus 5 measured
APEX-Agents65.8%43.0%Claude Opus 5
OSWorld 2.031.4%Not measuredOnly Claude Opus 5 measured
BALROG63.4%45.6%Claude Opus 5
PostTrainBench35.0%Not measuredOnly Claude Opus 5 measured
Vending-Bench 2$11,182$4,095Claude Opus 5
LMArena Search1242Not measuredOnly Claude Opus 5 measured
Neither model has a result on the other 14 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index160
Reasoning benchmarks for Claude Opus 5 and GPT-5.6 Luna
BenchmarkClaude Opus 5GPT-5.6 LunaLeader
GPQA Diamond93.4%91.7%Claude Opus 5
LiveBench reasoning91.2%85.6%Claude Opus 5
MysteryMechanism (Vals AI)37.4%14.4%Claude Opus 5
MMMU Pro (Vals AI)89.9%85.0%Claude Opus 5
GPQA Diamond (Artificial Analysis run)93.7%91.1%Claude Opus 5
Humanity's Last Exam (Artificial Analysis run)54.9%39.5%Claude Opus 5
GPQA Diamond (Epoch AI run)93.9%91.6%Claude Opus 5
ARC-AGI97.5%88.0%Claude Opus 5
ARC-AGI-290.4%59.5%Claude Opus 5
SimpleBench80.6%46.8%Claude Opus 5
Chess puzzles43.0%43.0%Tie
Mystery game puzzles59.0%21.0%Claude Opus 5
EBR-bench45.7%Not measuredOnly Claude Opus 5 measured
LMArena Text: hard prompts15271451Claude Opus 5
LMArena Text: expert15541478Claude Opus 5
LMArena Text: hard prompts in English15231460Claude Opus 5
LMArena Vision: diagrams13481273Claude Opus 5
LMArena Vision: homework13381282Claude Opus 5
Kagi LLM BenchmarkNot measured49.1%Only GPT-5.6 Luna measured
Neither model has a result on the other 3 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index60
Math benchmarks for Claude Opus 5 and GPT-5.6 Luna
BenchmarkClaude Opus 5GPT-5.6 LunaLeader
OTIS Mock AIME98.9%98.3%Claude Opus 5
FrontierMath, tiers 1 to 385.6%82.1%Claude Opus 5
FrontierMath, tier 473.2%61.0%Claude Opus 5
ProofBench99.0%60.0%Claude Opus 5
LMArena Text: math15331459Claude Opus 5
LMArena Text: mathematical work15441470Claude Opus 5
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index101
Knowledge benchmarks for Claude Opus 5 and GPT-5.6 Luna
BenchmarkClaude Opus 5GPT-5.6 LunaLeader
MMLU Pro91.6%86.0%Claude Opus 5
LegalBench (Vals AI)87.0%84.0%Claude Opus 5
CorpFin (Vals AI)73.2%64.2%Claude Opus 5
TaxEval (Vals AI)75.1%76.2%GPT-5.6 Luna
MortgageTax (Vals AI)72.1%67.3%Claude Opus 5
MedCode (Vals AI)63.6%42.4%Claude Opus 5
SimpleQA Verified59.9%41.0%Claude Opus 5
LMArena Text: science15311433Claude Opus 5
LMArena Text: medicine and healthcare15251399Claude Opus 5
LMArena Text: legal and government15301446Claude Opus 5
LMArena Text: business and finance15011439Claude Opus 5
FACTS Benchmark SuiteNot measured35.0%Only GPT-5.6 Luna measured
Neither model has a result on the other 6 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index90
Web design benchmarks for Claude Opus 5 and GPT-5.6 Luna
BenchmarkClaude Opus 5GPT-5.6 LunaLeader
Design Arena: websites13091255Claude Opus 5
Design Arena: UI components13461248Claude Opus 5
Design Arena: game dev13371194Claude Opus 5
Design Arena: data viz13291211Claude Opus 5
Design Arena: 3D13341121Claude Opus 5
Design Arena: all code categories13231231Claude Opus 5
Design Arena: SVG13111194Claude Opus 5
Design Arena: ASCII art13521142Claude Opus 5
LiveBench design83.0%76.0%Claude Opus 5
GeneralLike-for-likePublic boards, shown as published, kept out of the index312
General benchmarks for Claude Opus 5 and GPT-5.6 Luna
BenchmarkClaude Opus 5GPT-5.6 LunaLeader
LiveBench data analysis74.6%78.0%GPT-5.6 Luna
Public Benefits Bench (Vals AI)76.9%61.2%Claude Opus 5
SAGE (Vals AI)49.4%44.2%Claude Opus 5
MedScribe (Vals AI)91.0%84.4%Claude Opus 5
Vals Multimodal Index73.9%69.1%Claude Opus 5
Long context reasoning (AA-LCR)82.0%83.7%GPT-5.6 Luna
DTBench97.9%89.1%Claude Opus 5
LMCA64.5%48.5%Claude Opus 5
Surface Evolver BenchNot measured61.9%Only GPT-5.6 Luna measured
Furniture assembly60.8%42.5%Claude Opus 5
LMArena Text15091431Claude Opus 5
LMArena Text: instruction following15181437Claude Opus 5
LMArena Text: creative writing14931396Claude Opus 5
LMArena Vision13201257Claude Opus 5
LMArena Text: English15081441Claude Opus 5
LMArena Text: non-English15021417Claude Opus 5
LMArena Text: Chinese15731470Claude Opus 5
LMArena Text: German15271455Claude Opus 5
LMArena Text: Japanese15201415Claude Opus 5
LMArena Text: Korean15171415Claude Opus 5
LMArena Text: French15191458Claude Opus 5
LMArena Text: Spanish15231446Claude Opus 5
LMArena Text: Russian15071427Claude Opus 5
LMArena Text: Polish15221405Claude Opus 5
LMArena Text: multi-turn15001434Claude Opus 5
LMArena Text: longer queries15171436Claude Opus 5
LMArena Text: writing and language14991419Claude Opus 5
LMArena Text: entertainment and media14811402Claude Opus 5
LMArena Vision: English13151263Claude Opus 5
LMArena Vision: Chinese14001294Claude Opus 5
LMArena Vision: reading text13291264Claude Opus 5
LMArena Vision: humor12931190Claude Opus 5
LMArena Vision: creative writing13131257Claude Opus 5
LMArena Document15161457Claude Opus 5
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 Claude Opus 5 and GPT-5.6 Luna
BenchmarkClaude Opus 5GPT-5.6 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

Claude Opus 5$0.018Fits in one request
GPT-5.6 Luna$0.0008Fits in one request

GPT-5.6 Luna costs less

Repository review

50K input + 3K output tokens

Claude Opus 5$0.325Fits in one request
GPT-5.6 Luna$0.014Fits in one request

GPT-5.6 Luna costs less

Agent loop

200K input + 10K output tokens

Claude Opus 5$1.25Fits in one request
GPT-5.6 Luna$0.052Fits in one request

GPT-5.6 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 Claude Opus 5 and GPT-5.6 Luna
SpecificationAnthropic Claude Opus 5OpenAI GPT-5.6 Luna
MakerAnthropicOpenAI
WeightsNo public weights listedNo public weights listed
ReleasedJuly 24, 2026July 9, 2026
Context window1M1.05M
Input $/Mtok$5.00$0.20
Output $/Mtok$25.00$1.20
Coding index83.058.0
Independent measurements behind it66
Cost per solved task$1.33$0.04
SourceOpenRouter listingOpenRouter listing

Questions

Is Claude Opus 5 better than GPT-5.6 Luna for coding?

Pick Claude Opus 5 for coding: it scores 83.0 on the Vibe Coding Index against 58.0 for GPT-5.6 Luna. Pick GPT-5.6 Luna for output tokens 20.8x cheaper ($1.20 vs $25.00 per million) and a lower cost per solved task ($0.04 against $1.33).

Which is cheaper, Claude Opus 5 or GPT-5.6 Luna?

Claude Opus 5 costs $5.00 per million input tokens and $25.00 per million output tokens; GPT-5.6 Luna costs $0.20 and $1.20. Per solved SWE-bench Verified task, Claude Opus 5 costs $1.33 and GPT-5.6 Luna $0.04.

Which has the bigger context window, Claude Opus 5 or GPT-5.6 Luna?

GPT-5.6 Luna, with 1.05M tokens against 1M.

How many benchmarks is the Claude Opus 5 vs GPT-5.6 Luna comparison based on?

116 benchmarks have a result for both models as of October 22, 2026. Claude Opus 5's coding index rests on 6 independent benchmarks and GPT-5.6 Luna's on 6. 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/claude-opus-5-vs-gpt-5-6-luna" title="Claude Opus 5 vs GPT-5.6 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/claude-opus-5-vs-gpt-5-6-luna">LLMBenchmark.ai, Claude Opus 5 vs GPT-5.6 Luna</a></p>