Gemini 3 Pro vs Gemma 4 31B
Wins 5 of 8 areas
Coding · Reasoning · Images and charts · Math · Long documents
Wins 1 of 8 areas
Following instructions
Gemini 3 Pro is the stronger all-rounder.Gemma 4 31B is cheaper and better at following instructions.
Scores updated · 45 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 software70Gemini 3 Pro7 of 7 tests
- Images and chartsUnderstanding pictures, charts and video40Gemini 3 Pro4 of 4 tests
- ReasoningHard problems that need careful thinking30Gemini 3 Pro3 of 3 tests
- MathCompetition and research-level math30Gemini 3 Pro3 of 3 tests
- Long documentsFinding answers in very long texts20Gemini 3 Pro2 of 2 tests
- Following instructionsDoing exactly what it is asked01Gemma 4 31B1 of 1 test
- AgentsCarrying out multi-step tasks on its own11Even1 each
- FactsGetting facts right instead of making them up22Even2 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.
Gemma 4 31B costs 96% 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
- Short factual questions, answered correctlySimpleQA Verified+62.5points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+35.8points ahead
- Real work tasks from 44 professionsGDPVal+28.1points ahead
Where Gemma 4 31B pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+6.5points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+5.2points ahead
- Multi-step tasks using many real software toolsMCP Atlas+3.1points ahead
Every test, side by side
All 45 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 3 Pro
- SWE-bench VerifiedGemini 3 Pro by 24.276.252+24.2
- SWE-bench MultilingualGemini 3 Pro by 13.36551.7+13.3
- SciCodeGemini 3 Pro by 10.656.145.5+10.6
- SWE-bench ProGemini 3 Pro by 7.643.335.7+7.6
- LMArena · WebDevGemini 3 Pro by 75 rating points14401365+75 rating
- LiveCodeBench v6Gemini 3 Pro by 7.487.480+7.4
- Terminal-Bench HardGemini 3 Pro by 5.341.736.4+5.3
AgentsEven
ReasoningGemini 3 Pro
- Humanity's Last ExamGemini 3 Pro by 16.139.723.6+16.1
- CritPtGemini 3 Pro by 7.79.11.4+7.7
- GPQA DiamondGemini 3 Pro by 5.190.885.7+5.1
FactsEven
- SimpleQA VerifiedGemini 3 Pro by 62.572.910.4+62.5
- AA-Omniscience · AccuracyGemini 3 Pro by 35.855.820+35.8
- AA-Omniscience · Non-hallucinationGemma 4 31B by 6.58.515+6.5
- Vectara HHEM hallucination ratelower is betterGemma 4 31B by 6.213.67.4+6.2
Images and chartsGemini 3 Pro
- CharXiv (RQ)Gemini 3 Pro by 13.581.467.9+13.5
- MathVistaGemini 3 Pro by 10.589.879.3+10.5
- MMMU-ProGemini 3 Pro by 6.880.273.4+6.8
- LMArena · VisionGemini 3 Pro by 28 rating points13051277+28 rating
MathGemini 3 Pro
- HMMT Feb. 2026Gemini 3 Pro by 9.286.477.2+9.2
- IMOAnswerBenchGemini 3 Pro by 8.683.174.5+8.6
- AIME 2026Gemini 3 Pro by 2.591.789.2+2.5
Long documentsGemini 3 Pro
- MRCR v2 8 needle 128k (average)Gemini 3 Pro by 10.67766.4+10.6
- AA-LCRGemini 3 Pro by 6.37669.7+6.3
Following instructionsGemma 4 31B
- IFBenchGemma 4 31B by 5.270.475.6+5.2
Other results19 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 Pro by 63.215.3-47.9+63.2
- AA Agentic IndexGemini 3 Pro by 45.3526.7+45.3
- τ²-Bench Telecom (AA run)Gemini 3 Pro by 27.287.159.9+27.2
- τ²-BenchGemini 3 Pro by 21.19876.9+21.1
- HLE (with tools)Gemini 3 Pro by 19.345.826.5+19.3
- SimpleVQAGemini 3 Pro by 16.869.752.9+16.8
- AA IntelligenceGemini 3 Pro by 13.32814.7+13.3
- Terminal-Bench 2.0Gemini 3 Pro by 11.354.242.9+11.3
- HMMT Feb. 2025Gemini 3 Pro by 8.897.588.7+8.8
- OmniDocBench 1.5Gemini 3 Pro by 8.488.580.1+8.4
- vectara_factual_consistencyGemma 4 31B by 6.286.492.6+6.2
- HMMT Nov. 2025Gemini 3 Pro by 5.893.387.5+5.8
- MMLU-ProGemini 3 Pro by 4.990.185.2+4.9
- MMMLUGemini 3 Pro by 3.491.888.4+3.4
- Artificial Analysis Coding IndexGemini 3 Pro by 3.146.543.4+3.1
- vectara_avg_summary_lengthGemini 3 Pro by 26.1 rating points101.975.8+26.1 rating
- t2-benchGemma 4 31B by 185.486.4+1
- vectara_answer_ratetie99.4100tie
- MATH-Visiontie86.185.6tie
Questions people ask
Which is better, Gemini 3 Pro or Gemma 4 31B?
Gemini 3 Pro wins five of the eight areas we test: coding, reasoning, images and charts, math and long documents. Gemma 4 31B wins following instructions, and costs 96% less. They are level on agents and facts.
Which is better for coding?
Gemini 3 Pro. It wins 7 of the 7 coding tests both models report; Gemma 4 31B wins none.
Which is cheaper?
Gemini 3 Pro costs $2.00 per million input tokens and $12.00 per million output tokens; Gemma 4 31B costs $0.17 and $0.40. That makes Gemma 4 31B about 96% cheaper for the same work.
How do you compare the two?
We use the 45 benchmark tests both models have published scores on. The verdict counts the 26 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 19 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.