Gemini 3.5 Flash vs GPT-5.4 mini
Wins 7 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math · Following instructions
Wins 1 of 8 areas
Long documents
Gemini 3.5 Flash is the stronger all-rounder.GPT-5.4 mini is cheaper and better at long documents.
Scores updated · 43 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.
- AgentsCarrying out multi-step tasks on its own70Gemini 3.5 Flash7 of 7 tests
- ReasoningHard problems that need careful thinking50Gemini 3.5 Flash5 of 5 tests
- CodingWriting and fixing software51Gemini 3.5 Flash5 of 7 tests · 1 tie
- FactsGetting facts right instead of making them up30Gemini 3.5 Flash3 of 3 tests
- MathCompetition and research-level math30Gemini 3.5 Flash3 of 3 tests
- Images and chartsUnderstanding pictures, charts and video20Gemini 3.5 Flash2 of 2 tests
- Following instructionsDoing exactly what it is asked20Gemini 3.5 Flash2 of 2 tests
- Long documentsFinding answers in very long texts01GPT-5.4 mini1 of 1 test
Long documents 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-5.4 mini costs 50% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Gemini 3.5 Flash pulls ahead
- Abstract visual puzzles that people can solveARC-AGI-2+53.2points ahead
- Short factual questions, answered correctlySimpleQA Verified+36.8points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+28points ahead
Where GPT-5.4 mini pulls ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+11.4points ahead
- Reasons across sets of long documentsAA-LCR+3.7points ahead
Every test, side by side
All 43 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 3.5 Flash
- Terminal-Bench 2.1Gemini 3.5 Flash by 19.578.759.2+19.5
- Terminal-Bench HardGPT-5.4 mini by 11.440.952.3+11.4
- LMArena · WebDevGemini 3.5 Flash by 102 rating points14991397+102 rating
- LiveBench · Agentic CodingGemini 3.5 Flash by 7.34941.7+7.3
- LiveBench · CodingGemini 3.5 Flash by 6.678.271.6+6.6
- SciCodeGemini 3.5 Flash by 1.853.952.1+1.8
- SWE-bench Protie55.154.4tie
AgentsGemini 3.5 Flash
- MCP AtlasGemini 3.5 Flash by 25.983.657.7+25.9
- AA ApexAgentsGemini 3.5 Flash by 18.947.128.2+18.9
- GDPValGemini 3.5 Flash by 9.435.225.8+9.4
- τ-Bench V3 · BankingGemini 3.5 Flash by 6.632.225.6+6.6
- OSWorld-VerifiedGemini 3.5 Flash by 6.378.472.1+6.3
- AA IT-Bench SREGemini 3.5 Flash by 5.140.335.2+5.1
- Terminal-Bench 4.0Gemini 3.5 Flash by 4.66.62+4.6
ReasoningGemini 3.5 Flash
- ARC-AGI-2Gemini 3.5 Flash by 53.272.118.9+53.2
- Humanity's Last ExamGemini 3.5 Flash by 14.642.728.1+14.6
- LiveBench · ReasoningGemini 3.5 Flash by 10.78271.3+10.7
- GPQA DiamondGemini 3.5 Flash by 4.792.287.5+4.7
- CritPtGemini 3.5 Flash by 3.113.110+3.1
FactsGemini 3.5 Flash
- SimpleQA VerifiedGemini 3.5 Flash by 36.866.229.4+36.8
- AA-Omniscience · Non-hallucinationGemini 3.5 Flash by 2837.89.8+28
- AA-Omniscience · AccuracyGemini 3.5 Flash by 13.951.437.5+13.9
Images and chartsGemini 3.5 Flash
- MMMU-ProGemini 3.5 Flash by 1184.373.3+11
- LMArena · VisionGemini 3.5 Flash by 65 rating points13101245+65 rating
MathGemini 3.5 Flash
- FrontierMath Tier 4Gemini 3.5 Flash by 1726.89.8+17
- FrontierMath Tiers 1-3 (v2)Gemini 3.5 Flash by 11.662.851.2+11.6
- LiveBench · MathematicsGemini 3.5 Flash by 9.788.278.5+9.7
Following instructionsGemini 3.5 Flash
- LiveBench · Instruction FollowingGemini 3.5 Flash by 15.875.659.8+15.8
- IFBenchGemini 3.5 Flash by 376.373.3+3
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.
- AA-OmniscienceGemini 3.5 Flash by 40.121.2-18.9+40.1
- ARC-AGI-1Gemini 3.5 Flash by 28.892.563.7+28.8
- Terminal-Bench 2.0Gemini 3.5 Flash by 16.276.260+16.2
- Artificial Analysis Coding IndexGemini 3.5 Flash by 1470.156.1+14
- livebench_languageGemini 3.5 Flash by 13.684.671+13.6
- ToolathlonGemini 3.5 Flash by 13.656.542.9+13.6
- Finance Agent v2Gemini 3.5 Flash by 12.557.945.4+12.5
- frontiermath_tier_4_v1Gemini 3.5 Flash by 12.514.62.1+12.5
- τ²-Bench Telecom (AA run)Gemini 3.5 Flash by 1295.383.3+12
- AA IntelligenceGemini 3.5 Flash by 8.532.624.1+8.5
- AA Agentic IndexGemini 3.5 Flash by 7.627.319.7+7.6
- livebench_data_analysisGPT-5.4 mini by 5.964.970.8+5.9
- Legal Agent Benchmarktie0.80tie
Questions people ask
Which is better, Gemini 3.5 Flash or GPT-5.4 mini?
Gemini 3.5 Flash wins seven of the eight areas we test: coding, agents, reasoning, facts, images and charts, math and following instructions. GPT-5.4 mini wins long documents, and costs 50% less.
Which is better for coding?
Gemini 3.5 Flash. It wins 5 of the 7 coding tests both models report; GPT-5.4 mini wins 1, and 1 is a tie.
Which is cheaper?
Gemini 3.5 Flash costs $1.50 per million input tokens and $9.00 per million output tokens; GPT-5.4 mini costs $0.75 and $4.50. That makes GPT-5.4 mini about 50% cheaper for the same work.
How do you compare the two?
We use the 43 benchmark tests both models have published scores on. The verdict counts the 30 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.