Gemini 3.5 Flash vs GPT-5.6 Luna
Wins 5 of 8 areas
Agents · Reasoning · Facts · Images and charts · Following instructions
Wins 3 of 8 areas
Coding · Math · Long documents
Gemini 3.5 Flash is the stronger all-rounder.GPT-5.6 Luna is cheaper and better at coding.
Scores updated · 47 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.
- Images and chartsUnderstanding pictures, charts and video40Gemini 3.5 Flash4 of 4 tests
- FactsGetting facts right instead of making them up30Gemini 3.5 Flash3 of 3 tests
- ReasoningHard problems that need careful thinking42Gemini 3.5 Flash4 of 6 tests
- AgentsCarrying out multi-step tasks on its own32Gemini 3.5 Flash3 of 6 tests · 1 tie
- Following instructionsDoing exactly what it is asked10Gemini 3.5 Flash1 of 1 test
- CodingWriting and fixing software04GPT-5.6 Luna4 of 6 tests · 2 ties
- MathCompetition and research-level math14GPT-5.6 Luna4 of 5 tests
- Long documentsFinding answers in very long texts02GPT-5.6 Luna2 of 2 tests
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 87% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Gemini 3.5 Flash pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+30.4points ahead
- Common-sense trick questionsSimpleBench+29.9points ahead
- Short factual questions, answered correctlySimpleQA Verified+25.2points ahead
Where GPT-5.6 Luna pulls ahead
- Finds one of eight look-alike replies in a long chatMRCR v2 8 needle 128k (average)+48.2points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+31.7points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+19.3points ahead
Every test, side by side
All 47 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-5.6 Luna
- SWE-bench ProGPT-5.6 Luna by 7.655.162.7+7.6
- LiveBench · CodingGPT-5.6 Luna by 4.778.282.9+4.7
- Terminal-Bench 2.1GPT-5.6 Luna by 2.278.780.9+2.2
- LMArena · WebDevGPT-5.6 Luna by 19 rating points15001519+19 rating
- LiveBench · Agentic Codingtie4948.4tie
- SciCodetie53.953.6tie
AgentsGemini 3.5 Flash
- GDPValGPT-5.6 Luna by 12.935.148+12.9
- AA ApexAgentsGemini 3.5 Flash by 11.347.135.8+11.3
- OSWorld-VerifiedGemini 3.5 Flash by 5.878.472.6+5.8
- Terminal-Bench 4.0GPT-5.6 Luna by 56.611.6+5
- τ-Bench V3 · BankingGemini 3.5 Flash by 1.132.231.1+1.1
- AA IT-Bench SREtie40.340.3tie
ReasoningGemini 3.5 Flash
- SimpleBenchGemini 3.5 Flash by 29.976.746.8+29.9
- ARC-AGI-2Gemini 3.5 Flash by 12.572.159.6+12.5
- CritPtGPT-5.6 Luna by 7.513.120.6+7.5
- LiveBench · ReasoningGPT-5.6 Luna by 3.68285.6+3.6
- Humanity's Last ExamGemini 3.5 Flash by 3.242.739.5+3.2
- GPQA DiamondGemini 3.5 Flash by 1.192.291.1+1.1
FactsGemini 3.5 Flash
- AA-Omniscience · Non-hallucinationGemini 3.5 Flash by 30.437.87.4+30.4
- SimpleQA VerifiedGemini 3.5 Flash by 25.266.241+25.2
- AA-Omniscience · AccuracyGemini 3.5 Flash by 8.751.442.7+8.7
Images and chartsGemini 3.5 Flash
- MMMU-ProGemini 3.5 Flash by 5.784.378.6+5.7
- Video-MMEGemini 3.5 Flash by 5.687.281.6+5.6
- LMArena · VisionGemini 3.5 Flash by 52 rating points13081256+52 rating
- CharXiv (RQ)Gemini 3.5 Flash by 1.584.282.7+1.5
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 19.362.882.1+19.3
- HMMT Feb. 2026GPT-5.6 Luna by 395.598.5+3
- AIME 2026GPT-5.6 Luna by 2.69597.6+2.6
- LiveBench · MathematicsGemini 3.5 Flash by 188.287.2+1
Long documentsGPT-5.6 Luna
- MRCR v2 8 needle 128k (average)GPT-5.6 Luna by 48.226.674.8+48.2
- AA-LCRGPT-5.6 Luna by 10.473.383.7+10.4
Following instructionsGemini 3.5 Flash
- LiveBench · Instruction FollowingGemini 3.5 Flash by 15.575.660.1+15.5
Other results14 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.
- AA-OmniscienceGemini 3.5 Flash by 31.521.2-10.3+31.5
- DeepSWEGPT-5.6 Luna by 30.23767.2+30.2
- DeepSWE 1.1GPT-5.6 Luna by 303767+30
- GDPval-AA v2 EloGPT-5.6 Luna by 235 rating points13491584+235 rating
- GDPval-AA v2GPT-5.6 Luna by 181 rating points13491530+181 rating
- LVBenchGemini 3.5 Flash by 17.476.358.9+17.4
- AA Agentic IndexGPT-5.6 Luna by 15.427.342.7+15.4
- livebench_data_analysisGPT-5.6 Luna by 13.164.978+13.1
- livebench_languageGemini 3.5 Flash by 1284.672.6+12
- AA IntelligenceGPT-5.6 Luna by 4.732.637.3+4.7
- ToolathlonGemini 3.5 Flash by 3.156.553.4+3.1
- MLE-BenchGemini 3.5 Flash by 2.149.747.6+2.1
- ARC-AGI-1Gemini 3.5 Flash by 1.892.590.7+1.8
- Artificial Analysis Coding IndexGPT-5.6 Luna by 1.370.171.4+1.3
Questions people ask
Which is better, Gemini 3.5 Flash or GPT-5.6 Luna?
Gemini 3.5 Flash wins five of the eight areas we test: agents, reasoning, facts, images and charts and following instructions. GPT-5.6 Luna wins coding, math and long documents, and costs 87% less.
Which is better for coding?
GPT-5.6 Luna. It wins 4 of the 6 coding tests both models report; Gemini 3.5 Flash wins none, and 2 are ties.
Which is cheaper?
Gemini 3.5 Flash costs $1.50 per million input tokens and $9.00 per million output tokens; GPT-5.6 Luna costs $0.20 and $1.20. That makes GPT-5.6 Luna about 87% cheaper for the same work.
How do you compare the two?
We use the 47 benchmark tests both models have published scores on. The verdict counts the 33 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 14 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.