Claude Opus 4.6 vs GPT-5.6 Luna
Wins 3 of 8 areas
Reasoning · Facts · Following instructions
Wins 3 of 8 areas
Coding · Images and charts · Math
The two are evenly matched.Claude Opus 4.6 is better at facts and reasoning; GPT-5.6 Luna at coding and math.
Scores updated · 44 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- CodingWriting and fixing software14GPT-5.6 Luna4 of 6 tests · 1 tie
- MathCompetition and research-level math13GPT-5.6 Luna3 of 5 tests · 1 tie
- Images and chartsUnderstanding pictures, charts and video12GPT-5.6 Luna2 of 3 tests
- FactsGetting facts right instead of making them up30Claude Opus 4.63 of 3 tests
- ReasoningHard problems that need careful thinking32Claude Opus 4.63 of 7 tests · 2 ties
- Following instructionsDoing exactly what it is asked10Claude Opus 4.61 of 1 test
- AgentsCarrying out multi-step tasks on its own11Even1 each · 2 ties
- Long documentsFinding answers in very long texts11Even1 each
Following instructions rests on a single test.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
GPT-5.6 Luna costs 95% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where GPT-5.6 Luna pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+31.7points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+16.1points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+12.2points ahead
Where Claude Opus 4.6 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+29.8points ahead
- Common-sense trick questionsSimpleBench+20.8points ahead
- Abstract visual puzzles that people can solveARC-AGI-2+9.2points ahead
Every test, side by side
All 44 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-5.6 Luna
- SWE-bench VerifiedGPT-5.6 Luna by 12.280.893+12.2
- SWE-bench ProGPT-5.6 Luna by 9.353.462.7+9.3
- LiveBench · CodingGPT-5.6 Luna by 4.778.282.9+4.7
- LMArena · WebDevClaude Opus 4.6 by 28 rating points15471519+28 rating
- SciCodeGPT-5.6 Luna by 1.751.953.6+1.7
- LiveBench · Agentic Codingtie4948.4tie
AgentsEven
- GDPValClaude Opus 4.6 by 7.955.948+7.9
- AA ApexAgentsGPT-5.6 Luna by 2.83335.8+2.8
- BrowseComptie8483.3tie
- OSWorld-Verifiedtie72.772.6tie
ReasoningClaude Opus 4.6
- SimpleBenchClaude Opus 4.6 by 20.867.646.8+20.8
- ARC-AGI-2Claude Opus 4.6 by 9.268.859.6+9.2
- CritPtGPT-5.6 Luna by 812.620.6+8
- LiveBench · ReasoningClaude Opus 4.6 by 3.188.785.6+3.1
- GPQA DiamondGPT-5.6 Luna by 1.589.691.1+1.5
- Humanity's Last Examtie39.939.5tie
- ARC-AGI-3tie0.50.2tie
FactsClaude Opus 4.6
- AA-Omniscience · Non-hallucinationClaude Opus 4.6 by 29.837.27.4+29.8
- SimpleQA VerifiedClaude Opus 4.6 by 64741+6
- AA-Omniscience · AccuracyClaude Opus 4.6 by 4.34742.7+4.3
Images and chartsGPT-5.6 Luna
- LMArena · VisionClaude Opus 4.6 by 60 rating points13161256+60 rating
- CharXiv (RQ)GPT-5.6 Luna by 5.377.482.7+5.3
- MMMU-ProGPT-5.6 Luna by 3.275.478.6+3.2
MathGPT-5.6 Luna
- FrontierMath Tier 4GPT-5.6 Luna by 31.726.858.5+31.7
- FrontierMath Tiers 1-3 (v2)GPT-5.6 Luna by 16.16682.1+16.1
- HMMT Feb. 2026GPT-5.6 Luna by 2.396.298.5+2.3
- LiveBench · MathematicsClaude Opus 4.6 by 2.189.387.2+2.1
- AIME 2026tie96.797.6tie
Long documentsEven
- MRCR v2 8 needle 128k (average)Claude Opus 4.6 by 9.28474.8+9.2
- AA-LCRGPT-5.6 Luna by 5.77883.7+5.7
Following instructionsClaude Opus 4.6
- LiveBench · Instruction FollowingClaude Opus 4.6 by 3.263.360.1+3.2
Other results13 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- Agents' Last ExamGPT-5.6 Luna by 25.125.250.3+25.1
- AA Agentic IndexClaude Opus 4.6 by 24.967.642.7+24.9
- AA-OmniscienceClaude Opus 4.6 by 2413.7-10.3+24
- Artificial Analysis Coding IndexGPT-5.6 Luna by 23.348.171.4+23.3
- ERQAGPT-5.6 Luna by 15.740.856.5+15.7
- livebench_languageClaude Opus 4.6 by 10.783.372.6+10.7
- livebench_data_analysisGPT-5.6 Luna by 8.169.978+8.1
- AA IntelligenceGPT-5.6 Luna by 5.431.937.3+5.4
- HLE (with tools)Claude Opus 4.6 by 4.453.348.9+4.4
- LVBenchClaude Opus 4.6 by 4.16358.9+4.1
- ToolathlonClaude Opus 4.6 by 3.456.853.4+3.4
- ARC-AGI-1Claude Opus 4.6 by 2.39390.7+2.3
- RealWorldQAClaude Opus 4.6 by 1.573.972.4+1.5
Questions people ask
Which is better, Claude Opus 4.6 or GPT-5.6 Luna?
Claude Opus 4.6 and GPT-5.6 Luna each win three of the eight areas we test. Claude Opus 4.6 wins reasoning, facts and following instructions; GPT-5.6 Luna wins coding, images and charts and math. GPT-5.6 Luna costs 95% less. They are level on agents and long documents.
Which is better for coding?
GPT-5.6 Luna. It wins 4 of the 6 coding tests both models report; Claude Opus 4.6 wins 1, and 1 is a tie.
Which is cheaper?
Claude Opus 4.6 costs $5.00 per million input tokens and $25.00 per million output tokens; GPT-5.6 Luna costs $0.20 and $1.20. That makes GPT-5.6 Luna about 95% cheaper for the same work.
How do you compare the two?
We use the 44 benchmark tests both models have published scores on. The verdict counts the 31 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 13 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.