Gemini 3.5 Flash vs GPT-5.2
Wins 6 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Following instructions
Wins 2 of 8 areas
Math · Long documents
Gemini 3.5 Flash is the stronger all-rounder.GPT-5.2 is better at math.
Scores updated · 48 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.
- ReasoningHard problems that need careful thinking51Gemini 3.5 Flash5 of 6 tests
- 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
- CodingWriting and fixing software42Gemini 3.5 Flash4 of 6 tests
- AgentsCarrying out multi-step tasks on its own31Gemini 3.5 Flash3 of 4 tests
- Following instructionsDoing exactly what it is asked10Gemini 3.5 Flash1 of 2 tests · 1 tie
- MathCompetition and research-level math05GPT-5.25 of 5 tests
- Long documentsFinding answers in very long texts02GPT-5.22 of 2 tests
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.
Gemini 3.5 Flash costs 33% 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
- Common-sense trick questionsSimpleBench+30.9points ahead
- Short factual questions, answered correctlySimpleQA Verified+29.1points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+25.2points ahead
Where GPT-5.2 pulls ahead
- Finds one of eight look-alike replies in a long chatMRCR v2 8 needle 128k (average)+57.2points ahead
- Real work tasks from 44 professionsGDPVal+13.1points ahead
- Reasons across sets of long documentsAA-LCR+9.4points ahead
Every test, side by side
All 48 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 3.5 Flash
- SWE-bench ProGemini 3.5 Flash by 25.255.129.9+25.2
- LMArena · WebDevGemini 3.5 Flash by 84 rating points14991415+84 rating
- Terminal-Bench HardGPT-5.2 by 6.140.947+6.1
- LiveBench · CodingGemini 3.5 Flash by 2.178.276.1+2.1
- SciCodeGemini 3.5 Flash by 1.853.952.1+1.8
- LiveBench · Agentic CodingGPT-5.2 by 1.34950.3+1.3
AgentsGemini 3.5 Flash
- AA ApexAgentsGemini 3.5 Flash by 24.147.123+24.1
- MCP AtlasGemini 3.5 Flash by 2383.660.6+23
- τ-Bench V3 · BankingGemini 3.5 Flash by 21.132.211.1+21.1
- GDPValGPT-5.2 by 13.135.248.3+13.1
ReasoningGemini 3.5 Flash
- SimpleBenchGemini 3.5 Flash by 30.976.745.8+30.9
- ARC-AGI-2Gemini 3.5 Flash by 19.272.152.9+19.2
- Humanity's Last ExamGemini 3.5 Flash by 542.737.7+5
- GPQA DiamondGemini 3.5 Flash by 1.992.290.3+1.9
- CritPtGemini 3.5 Flash by 1.513.111.6+1.5
- LiveBench · ReasoningGPT-5.2 by 1.28283.2+1.2
FactsGemini 3.5 Flash
- SimpleQA VerifiedGemini 3.5 Flash by 29.166.237.1+29.1
- AA-Omniscience · Non-hallucinationGemini 3.5 Flash by 1937.818.8+19
- AA-Omniscience · AccuracyGemini 3.5 Flash by 7.151.444.3+7.1
Images and chartsGemini 3.5 Flash
- MMMU-ProGemini 3.5 Flash by 9.784.374.6+9.7
- LMArena · VisionGemini 3.5 Flash by 64 rating points13101246+64 rating
- CharXiv (RQ)Gemini 3.5 Flash by 2.184.282.1+2.1
- Video-MMEGemini 3.5 Flash by 1.287.286+1.2
MathGPT-5.2
- LiveBench · MathematicsGPT-5.2 by 588.293.2+5
- FrontierMath Tier 4GPT-5.2 by 4.926.831.7+4.9
- FrontierMath Tiers 1-3 (v2)GPT-5.2 by 4.662.867.4+4.6
- AIME 2026GPT-5.2 by 3.39598.3+3.3
- HMMT Feb. 2026GPT-5.2 by 1.595.597+1.5
Long documentsGPT-5.2
- MRCR v2 8 needle 128k (average)GPT-5.2 by 57.226.683.8+57.2
- AA-LCRGPT-5.2 by 9.473.382.7+9.4
Following instructionsGemini 3.5 Flash
- LiveBench · Instruction FollowingGemini 3.5 Flash by 13.875.661.8+13.8
- IFBenchtie76.375.4tie
Other results16 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 Agentic IndexGPT-5.2 by 32.927.360.2+32.9
- Terminal-Bench 2.0Gemini 3.5 Flash by 22.276.254+22.2
- AA-OmniscienceGemini 3.5 Flash by 22.121.2-0.9+22.1
- Artificial Analysis Coding IndexGemini 3.5 Flash by 21.470.148.7+21.4
- GDPval-AA (Elo)Gemini 3.5 Flash by 194 rating points16561462+194 rating
- livebench_data_analysisGPT-5.2 by 13.364.978.2+13.3
- τ²-Bench Telecom (AA run)Gemini 3.5 Flash by 10.595.384.8+10.5
- ToolathlonGemini 3.5 Flash by 10.256.546.3+10.2
- ARC-AGI-1Gemini 3.5 Flash by 6.392.586.2+6.3
- MotionBenchGemini 3.5 Flash by 5.870.664.8+5.8
- livebench_languageGemini 3.5 Flash by 4.884.679.8+4.8
- frontiermath_tier_4_v1GPT-5.2 by 4.214.618.8+4.2
- matharena_visual_math_overallGemini 3.5 Flash by 3.489.986.5+3.4
- AA IntelligenceGemini 3.5 Flash by 2.232.630.4+2.2
- LongVideoBenchGemini 3.5 Flash by 1.177.676.5+1.1
- LiveBenchtie7574.8tie
Questions people ask
Which is better, Gemini 3.5 Flash or GPT-5.2?
Gemini 3.5 Flash wins six of the eight areas we test: coding, agents, reasoning, facts, images and charts and following instructions. GPT-5.2 wins math and long documents.
Which is better for coding?
Gemini 3.5 Flash. It wins 4 of the 6 coding tests both models report; GPT-5.2 wins 2.
Which is cheaper?
Gemini 3.5 Flash costs $1.50 per million input tokens and $9.00 per million output tokens; GPT-5.2 costs $1.75 and $14.00. That makes Gemini 3.5 Flash about 33% cheaper for the same work.
How do you compare the two?
We use the 48 benchmark tests both models have published scores on. The verdict counts the 32 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 16 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.