Gemma 4 12B vs Gemma 4 26B A4B
Wins 1 of 8 areas
Following instructions
Wins 5 of 8 areas
Coding · Agents · Reasoning · Math · Long documents
Gemma 4 26B A4B is the stronger all-rounder.Gemma 4 12B is cheaper and better at following instructions.
Scores updated · 30 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 software13Gemma 4 26B A4B3 of 4 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 texts01Gemma 4 26B A4B1 of 2 tests · 1 tie
- MathCompetition and research-level math01Gemma 4 26B A4B1 of 1 test
- Following instructionsDoing exactly what it is asked10Gemma 4 12B1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
- Images and chartsUnderstanding pictures, charts and video00Even0 each · 1 tie
Images and charts, 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 4 12B costs 25% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Gemma 4 26B A4B pulls ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+11.7points ahead
- US invitational high-school math exam problemsAIME 2026+10.8points ahead
- Recent programming contest problemsLiveCodeBench v6+5.1points ahead
Where Gemma 4 12B pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+5.4points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+4.6points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+1.1points ahead
Every test, side by side
All 30 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemma 4 26B A4B
- Terminal-Bench 2.1Gemma 4 26B A4B by 11.727.339+11.7
- LiveCodeBench v6Gemma 4 26B A4B by 5.17277.1+5.1
- Terminal-Bench HardGemma 4 12B by 4.618.213.6+4.6
- SciCodeGemma 4 26B A4B by 1.838.240+1.8
AgentsGemma 4 26B A4B
- GDPValGemma 4 26B A4B by 3.403.4+3.4
- τ-Bench V3 · BankingGemma 4 26B A4B by 3.38.712+3.3
ReasoningGemma 4 26B A4B
- GPQA DiamondGemma 4 26B A4B by 3.975.379.2+3.9
- Humanity's Last ExamGemma 4 26B A4B by 3.615.719.3+3.6
- CritPttie00tie
FactsEven
- AA-Omniscience · Non-hallucinationGemma 4 12B by 5.41913.6+5.4
- AA-Omniscience · AccuracyGemma 4 26B A4B by 3.515.619.1+3.5
Long documentsGemma 4 26B A4B
- AA-LCRGemma 4 26B A4B by 263.765.7+2
- MRCR v2 8 needle 128k (average)tie43.444.1tie
Following instructionsGemma 4 12B
- IFBenchGemma 4 12B by 1.173.572.4+1.1
Other results14 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.
- HLE (with tools)Gemma 4 26B A4B by 125.217.2+12
- BIG-Bench Extra HardGemma 4 26B A4B by 11.85364.8+11.8
- MedXPertQA MMGemma 4 26B A4B by 9.448.758.1+9.4
- Artificial Analysis Coding IndexGemma 4 26B A4B by 8.33139.3+8.3
- τ²-Bench Telecom (AA run)Gemma 4 26B A4B by 7.336.343.6+7.3
- Codeforces ELOGemma 4 26B A4B by 59 rating points16591718+59 rating
- MMLU-ProGemma 4 26B A4B by 5.477.282.6+5.4
- AA Agentic IndexGemma 4 26B A4B by 3.17.911+3.1
- MMMLUGemma 4 26B A4B by 2.983.486.3+2.9
- MATH-VisionGemma 4 26B A4B by 2.779.782.4+2.7
- AA IntelligenceGemma 4 26B A4B by 2.514.216.7+2.5
- AA-OmniscienceGemma 4 26B A4B by 1.9-52.7-50.8+1.9
- τ²-Benchtie6968.2tie
- OmniDocBench 1.5 (average edit distance, lower is better)lower is bettertie0.20.1tie
Questions people ask
Which is better, Gemma 4 12B or Gemma 4 26B A4B?
Gemma 4 26B A4B wins five of the eight areas we test: coding, agents, reasoning, math and long documents. Gemma 4 12B wins following instructions, and costs 25% less. They are level on facts and images and charts.
Which is better for coding?
Gemma 4 26B A4B. It wins 3 of the 4 coding tests both models report; Gemma 4 12B wins 1.
Which is cheaper?
Gemma 4 12B costs $0.10 per million input tokens and $0.30 per million output tokens; Gemma 4 26B A4B costs $0.13 and $0.40. That makes Gemma 4 12B about 25% cheaper for the same work.
How do you compare the two?
We use the 30 benchmark tests both models have published scores on. The verdict counts the 16 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 14 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.