Gemma 3 27B vs Gemma 4 26B A4B
Wins 0 of 8 areas
—
Wins 8 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math · Long documents · Following instructions
Gemma 4 26B A4B is the stronger all-rounder.Gemma 3 27B is cheaper.
Scores updated · 40 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 software05Gemma 4 26B A4B5 of 5 tests
- Images and chartsUnderstanding pictures, charts and video04Gemma 4 26B A4B4 of 4 tests
- FactsGetting facts right instead of making them up03Gemma 4 26B A4B3 of 3 tests
- ReasoningHard problems that need careful thinking02Gemma 4 26B A4B2 of 3 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own02Gemma 4 26B A4B2 of 2 tests
- Long documentsFinding answers in very long texts02Gemma 4 26B A4B2 of 2 tests
- MathCompetition and research-level math01Gemma 4 26B A4B1 of 1 test
- Following instructionsDoing exactly what it is asked01Gemma 4 26B A4B1 of 1 test
Math and following instructions rest on a single test each.
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 3 27B costs 55% 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 Gemma 4 26B A4B pulls ahead
- US invitational high-school math exam problemsAIME 2026+67.5points ahead
- Reasons across sets of long documentsAA-LCR+58.4points ahead
- Recent programming contest problemsLiveCodeBench v6+48points ahead
Where Gemma 3 27B pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 40 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemma 4 26B A4B
- LiveCodeBench v6Gemma 4 26B A4B by 4829.177.1+48
- Terminal-Bench 2.1Gemma 4 26B A4B by 34.54.539+34.5
- SciCodeGemma 4 26B A4B by 16.723.340+16.7
- Terminal-Bench HardGemma 4 26B A4B by 9.83.813.6+9.8
- SWE-bench ProGemma 4 26B A4B by 2.411.413.8+2.4
AgentsGemma 4 26B A4B
- τ-Bench V3 · BankingGemma 4 26B A4B by 11.20.812+11.2
- GDPValGemma 4 26B A4B by 3.403.4+3.4
ReasoningGemma 4 26B A4B
- GPQA DiamondGemma 4 26B A4B by 36.442.879.2+36.4
- Humanity's Last ExamGemma 4 26B A4B by 14.94.419.3+14.9
- CritPttie00tie
FactsGemma 4 26B A4B
- AA-Omniscience · AccuracyGemma 4 26B A4B by 6.112.919.1+6.1
- AA-Omniscience · Non-hallucinationGemma 4 26B A4B by 5.77.913.6+5.7
- Vectara HHEM hallucination ratelower is betterGemma 4 26B A4B by 2.27.45.2+2.2
Images and chartsGemma 4 26B A4B
- MMMU-ProGemma 4 26B A4B by 21.24869.2+21.2
- MMMUGemma 4 26B A4B by 13.564.978.4+13.5
- MathVistaGemma 4 26B A4B by 11.867.679.4+11.8
- LMArena · VisionGemma 4 26B A4B by 96 rating points11641260+96 rating
Long documentsGemma 4 26B A4B
- AA-LCRGemma 4 26B A4B by 58.47.365.7+58.4
- MRCR v2 8 needle 128k (average)Gemma 4 26B A4B by 30.613.544.1+30.6
Following instructionsGemma 4 26B A4B
- IFBenchGemma 4 26B A4B by 40.631.872.4+40.6
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.
- Codeforces ELOGemma 4 26B A4B by 1608 rating points1101718+1608 rating
- τ²-BenchGemma 4 26B A4B by 5216.268.2+52
- BIG-Bench Extra HardGemma 4 26B A4B by 45.519.364.8+45.5
- HumanEvalGemma 3 27B by 37.887.850+37.8
- MATH-VisionGemma 4 26B A4B by 36.44682.4+36.4
- τ²-Bench Telecom (AA run)Gemma 4 26B A4B by 33.110.543.6+33.1
- vectara_avg_summary_lengthGemma 3 27B by 29.396.467.1+29.3
- Artificial Analysis Coding IndexGemma 4 26B A4B by 29.210.139.3+29.2
- AA-OmniscienceGemma 4 26B A4B by 16.4-67.2-50.8+16.4
- MMMLUGemma 4 26B A4B by 15.670.786.3+15.6
- MMLU-ProGemma 4 26B A4B by 1567.682.6+15
- AA IntelligenceGemma 4 26B A4B by 11.84.916.7+11.8
- MBPPGemma 3 27B by 674.468.4+6
- AI2DGemma 4 26B A4B by 3.884.588.3+3.8
- vectara_factual_consistencyGemma 4 26B A4B by 2.292.694.8+2.2
- IFEvalGemma 4 26B A4B by 190.491.4+1
- vectara_answer_rateGemma 4 26B A4B by 198.899.8+1
- MMLUtie76.977.8tie
- OmniDocBench 1.5 (average edit distance, lower is better)lower is bettertie0.40.1tie
Questions people ask
Which is better, Gemma 3 27B or Gemma 4 26B A4B?
Gemma 4 26B A4B wins all eight areas we test: coding, agents, reasoning, facts, images and charts, math, long documents and following instructions. Gemma 3 27B wins none, but costs 55% less.
Which is better for coding?
Gemma 4 26B A4B. It wins 5 of the 5 coding tests both models report; Gemma 3 27B wins none.
Which is cheaper?
Gemma 3 27B costs $0.08 per million input tokens and $0.16 per million output tokens; Gemma 4 26B A4B costs $0.13 and $0.40. That makes Gemma 3 27B about 55% cheaper for the same work.
How do you compare the two?
We use the 40 benchmark tests both models have published scores on. The verdict counts the 21 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.