Gemini 3 Flash Preview vs Gemini 3 Pro
Wins 3 of 8 areas
Agents · Math · Following instructions
Wins 3 of 8 areas
Reasoning · Facts · Images and charts
The two are evenly matched.Gemini 3 Flash Preview is better at agents and math; Gemini 3 Pro at facts and reasoning.
Scores updated · 50 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.
- FactsGetting facts right instead of making them up03Gemini 3 Pro3 of 4 tests · 1 tie
- ReasoningHard problems that need careful thinking13Gemini 3 Pro3 of 5 tests · 1 tie
- Images and chartsUnderstanding pictures, charts and video02Gemini 3 Pro2 of 3 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own30Gemini 3 Flash Preview3 of 3 tests
- MathCompetition and research-level math20Gemini 3 Flash Preview2 of 2 tests
- Following instructionsDoing exactly what it is asked10Gemini 3 Flash Preview1 of 1 test
- CodingWriting and fixing software22Even2 each · 1 tie
- 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.
Gemini 3 Flash Preview costs 75% 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 Pro pulls ahead
- Common-sense trick questionsSimpleBench+15.3points ahead
- Finds one of eight look-alike replies in a long chatMRCR v2 8 needle 128k (average)+9.8points ahead
- Short factual questions, answered correctlySimpleQA Verified+6.1points ahead
Where Gemini 3 Flash Preview pulls ahead
- Long professional tasks in banking, consulting and lawAA ApexAgents+9.3points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+7.6points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+6.3points ahead
Every test, side by side
All 50 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingEven
- SWE-bench ProGemini 3 Flash Preview by 6.349.643.3+6.3
- SciCodeGemini 3 Pro by 5.550.656.1+5.5
- Terminal-Bench HardGemini 3 Pro by 3.138.641.7+3.1
- SWE-bench VerifiedGemini 3 Flash Preview by 1.87876.2+1.8
- LMArena · WebDevtie14391439tie
AgentsGemini 3 Flash Preview
- AA ApexAgentsGemini 3 Flash Preview by 9.327.718.4+9.3
- MCP AtlasGemini 3 Flash Preview by 3.357.454.1+3.3
- GDPValGemini 3 Flash Preview by 135.234.2+1
ReasoningGemini 3 Pro
- SimpleBenchGemini 3 Pro by 15.361.176.4+15.3
- Humanity's Last ExamGemini 3 Pro by 3.136.639.7+3.1
- ARC-AGI-2Gemini 3 Flash Preview by 2.533.631.1+2.5
- GPQA DiamondGemini 3 Pro by 189.890.8+1
- CritPttie8.69.1tie
FactsGemini 3 Pro
- SimpleQA VerifiedGemini 3 Pro by 6.166.872.9+6.1
- AA-Omniscience · AccuracyGemini 3 Pro by 2.453.455.8+2.4
- AA-Omniscience · Non-hallucinationGemini 3 Pro by 1.578.5+1.5
- Vectara HHEM hallucination ratelower is bettertie13.513.6tie
Images and chartsGemini 3 Pro
- LMArena · VisionGemini 3 Pro by 20 rating points12851305+20 rating
- CharXiv (RQ)Gemini 3 Pro by 1.180.381.4+1.1
- MMMU-Protie79.980.2tie
MathGemini 3 Flash Preview
- AIME 2026Gemini 3 Flash Preview by 596.791.7+5
- HMMT Feb. 2026Gemini 3 Flash Preview by 389.486.4+3
Long documentsEven
- MRCR v2 8 needle 128k (average)Gemini 3 Pro by 9.867.277+9.8
- AA-LCRGemini 3 Flash Preview by 27876+2
Following instructionsGemini 3 Flash Preview
- IFBenchGemini 3 Flash Preview by 7.67870.4+7.6
Other results25 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.
- Vending-Bench 2Gemini 3 Pro by 184316 rating points363500547816+184316 rating
- frontiermath_tier_4_v1Gemini 3 Pro by 14.64.218.8+14.6
- ToolathlonGemini 3 Flash Preview by 1349.436.4+13
- ARC-AGI-1Gemini 3 Flash Preview by 9.784.775+9.7
- τ²-Bench Telecom (AA run)Gemini 3 Pro by 6.780.487.1+6.7
- Terminal-Bench 2.0Gemini 3 Pro by 6.647.654.2+6.6
- AA-OmniscienceGemini 3 Pro by 5.210.115.3+5.2
- t2-benchGemini 3 Flash Preview by 4.890.285.4+4.8
- Artificial Analysis Coding IndexGemini 3 Pro by 3.942.646.5+3.9
- ScreenSpot-Pro (No tools)Gemini 3 Pro by 3.669.172.7+3.6
- SimpleQAGemini 3 Pro by 3.468.772.1+3.4
- AA Agentic IndexGemini 3 Pro by 2.349.752+2.3
- AA IntelligenceGemini 3 Pro by 1.726.328+1.7
- matharena_visual_math_overallGemini 3 Flash Preview by 1.685.884.2+1.6
- vectara_avg_summary_lengthGemini 3 Pro by 11.7 rating points90.2101.9+11.7 rating
- LiveBenchGemini 3 Pro by 172.473.4+1
- GDPval-AA (Elo)tie12041195tie
- VideoMMMUtie86.987.6tie
- vectara_answer_ratetie99.899.4tie
- AIME 2025tie99.7100tie
- vectara_factual_consistencytie86.586.4tie
- HMMT Feb. 2025tie97.597.5tie
- HMMT Nov. 2025tie93.393.3tie
- MMMLUtie91.891.8tie
- mrcr_v2_8needle_1m_pointwisetie26.326.3tie
Questions people ask
Which is better, Gemini 3 Flash Preview or Gemini 3 Pro?
Gemini 3 Flash Preview and Gemini 3 Pro each win three of the eight areas we test. Gemini 3 Flash Preview wins agents, math and following instructions; Gemini 3 Pro wins reasoning, facts and images and charts. Gemini 3 Flash Preview costs 75% less. They are level on coding and long documents.
Which is better for coding?
Neither. They win 2 coding tests each of the 5 both models report, and 1 is a tie.
Which is cheaper?
Gemini 3 Flash Preview costs $0.50 per million input tokens and $3.00 per million output tokens; Gemini 3 Pro costs $2.00 and $12.00. That makes Gemini 3 Flash Preview about 75% cheaper for the same work.
How do you compare the two?
We use the 50 benchmark tests both models have published scores on. The verdict counts the 25 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 25 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.