Gemma 3 12B vs Qwen3.5 35B A3B
Wins 0 of 7 areas
—
Wins 7 of 7 areas
Coding · Agents · Reasoning · Facts · Images and charts · Long documents · Following instructions
Qwen3.5 35B A3B is the stronger all-rounder.Gemma 3 12B is cheaper.
Scores updated · 29 tests both models report · How we compare
Where each one wins
Tests won in each of the seven areas where both have results. 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 software03Qwen3.5 35B A3B3 of 3 tests
- Images and chartsUnderstanding pictures, charts and video03Qwen3.5 35B A3B3 of 3 tests
- ReasoningHard problems that need careful thinking02Qwen3.5 35B A3B2 of 3 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own02Qwen3.5 35B A3B2 of 2 tests
- FactsGetting facts right instead of making them up12Qwen3.5 35B A3B2 of 3 tests
- Long documentsFinding answers in very long texts01Qwen3.5 35B A3B1 of 1 test
- Following instructionsDoing exactly what it is asked01Qwen3.5 35B A3B1 of 1 test
Long documents 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 12B costs 91% 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 Qwen3.5 35B A3B pulls ahead
- Reasons across sets of long documentsAA-LCR+63.7points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+49.6points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+35.8points ahead
Where Gemma 3 12B pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 29 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingQwen3.5 35B A3B
- Terminal-Bench 2.1Qwen3.5 35B A3B by 40.8040.8+40.8
- Terminal-Bench HardQwen3.5 35B A3B by 25.70.826.5+25.7
- SciCodeQwen3.5 35B A3B by 21.316.437.7+21.3
AgentsQwen3.5 35B A3B
- GDPValQwen3.5 35B A3B by 5.605.6+5.6
- τ-Bench V3 · BankingQwen3.5 35B A3B by 4.10.84.9+4.1
ReasoningQwen3.5 35B A3B
- GPQA DiamondQwen3.5 35B A3B by 49.634.984.5+49.6
- Humanity's Last ExamQwen3.5 35B A3B by 16.84.221+16.8
- CritPttie00.9tie
FactsQwen3.5 35B A3B
- AA-Omniscience · Non-hallucinationQwen3.5 35B A3B by 122.614.6+12
- AA-Omniscience · AccuracyQwen3.5 35B A3B by 9.810.320.1+9.8
- Vectara HHEM hallucination ratelower is betterGemma 3 12B by 6.14.410.5+6.1
Images and chartsQwen3.5 35B A3B
Following instructionsQwen3.5 35B A3B
- IFBenchQwen3.5 35B A3B by 35.836.772.5+35.8
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.
- τ²-Bench Telecom (AA run)Qwen3.5 35B A3B by 78.410.889.2+78.4
- MATH-VisionQwen3.5 35B A3B by 51.832.183.9+51.8
- Artificial Analysis Coding IndexQwen3.5 35B A3B by 31.25.837+31.2
- AA-OmniscienceQwen3.5 35B A3B by 29-77.1-48.1+29
- MMLU-ProQwen3.5 35B A3B by 24.760.685.3+24.7
- HumanEvalGemma 3 12B by 18.985.466.5+18.9
- AA IntelligenceQwen3.5 35B A3B by 15.53.819.3+15.5
- AI2DQwen3.5 35B A3B by 8.484.292.6+8.4
- vectara_factual_consistencyGemma 3 12B by 6.195.689.5+6.1
- vectara_avg_summary_lengthQwen3.5 35B A3B by 5.289.794.9+5.2
- IFEvalQwen3.5 35B A3B by 388.991.9+3
- vectara_answer_rateQwen3.5 35B A3B by 2.497.499.8+2.4
- MBPPGemma 3 12B by 2.27370.8+2.2
Questions people ask
Which is better, Gemma 3 12B or Qwen3.5 35B A3B?
Qwen3.5 35B A3B wins all seven areas where both have results: coding, agents, reasoning, facts, images and charts, long documents and following instructions. Gemma 3 12B wins none, but costs 91% less.
Which is better for coding?
Qwen3.5 35B A3B. It wins 3 of the 3 coding tests both models report; Gemma 3 12B wins none.
Which is cheaper?
Gemma 3 12B costs $0.05 per million input tokens and $0.15 per million output tokens; Qwen3.5 35B A3B costs $0.25 and $2.00. That makes Gemma 3 12B about 91% cheaper for the same work.
How do you compare the two?
We use the 29 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 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.