Gemma 4 E4B vs Qwen3.5 9B
Wins 0 of 8 areas
—
Wins 7 of 8 areas
Coding · Agents · Reasoning · Images and charts · Math · Long documents · Following instructions
Qwen3.5 9B is the stronger all-rounder.Gemma 4 E4B is cheaper.
Scores updated · 32 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 software04Qwen3.5 9B4 of 4 tests
- ReasoningHard problems that need careful thinking02Qwen3.5 9B2 of 3 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own01Qwen3.5 9B1 of 3 tests · 2 ties
- Images and chartsUnderstanding pictures, charts and video01Qwen3.5 9B1 of 1 test
- MathCompetition and research-level math01Qwen3.5 9B1 of 1 test
- Long documentsFinding answers in very long texts01Qwen3.5 9B1 of 1 test
- Following instructionsDoing exactly what it is asked01Qwen3.5 9B1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
Images and charts, math, 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 4 E4B costs 64% 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 9B pulls ahead
- US invitational high-school math exam problemsAIME 2026+50points ahead
- Reasons across sets of long documentsAA-LCR+38points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+27.3points ahead
Where Gemma 4 E4B pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+52.7points ahead
Every test, side by side
All 32 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingQwen3.5 9B
- Terminal-Bench 2.1Qwen3.5 9B by 27.31.929.2+27.3
- Terminal-Bench HardQwen3.5 9B by 15.98.324.2+15.9
- LiveCodeBench v6Qwen3.5 9B by 13.65265.6+13.6
- SciCodeQwen3.5 9B by 5.124.429.5+5.1
AgentsQwen3.5 9B
- τ-Bench V3 · BankingQwen3.5 9B by 1.65.47+1.6
- Terminal-Bench 4.0tie00.5tie
- GDPValtie00tie
ReasoningQwen3.5 9B
- GPQA DiamondQwen3.5 9B by 2357.680.6+23
- Humanity's Last ExamQwen3.5 9B by 11.13.814.9+11.1
- CritPttie0.60.3tie
FactsEven
- AA-Omniscience · Non-hallucinationGemma 4 E4B by 52.769.116.4+52.7
- AA-Omniscience · AccuracyQwen3.5 9B by 7.88.616.4+7.8
Following instructionsQwen3.5 9B
- IFBenchQwen3.5 9B by 22.544.266.7+22.5
Other results16 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 9B by 6620.886.8+66
- AA-OmniscienceGemma 4 E4B by 33.8-19.7-53.5+33.8
- AIME25 no toolsQwen3.5 9B by 21.834.356.1+21.8
- t2-benchQwen3.5 9B by 21.657.579.1+21.6
- BFCLv4Qwen3.5 9B by 20.14060.1+20.1
- Artificial Analysis Coding IndexQwen3.5 9B by 19.39.428.7+19.3
- PinchBenchQwen3.5 9B by 16.455.171.5+16.4
- MMLU-ProQwen3.5 9B by 13.169.482.5+13.1
- BrowseComp+ (OpenClaw)Qwen3.5 9B by 11.315.927.2+11.3
- Claw-Eval AvgQwen3.5 9B by 8.55866.5+8.5
- LiveCodeBenchV6 no toolsQwen3.5 9B by 6.163.869.9+6.1
- AA Agentic IndexQwen3.5 9B by 5.21.87+5.2
- MMMLUQwen3.5 9B by 4.676.681.2+4.6
- IFEvalQwen3.5 9B by 3.687.991.5+3.6
- AA IntelligenceQwen3.5 9B by 2.38.911.2+2.3
- τ³-Bench BankingQwen3.5 9B by 1.14.15.2+1.1
Questions people ask
Which is better, Gemma 4 E4B or Qwen3.5 9B?
Qwen3.5 9B wins seven of the eight areas we test: coding, agents, reasoning, images and charts, math, long documents and following instructions. Gemma 4 E4B wins none, but costs 64% less. They are level on facts.
Which is better for coding?
Qwen3.5 9B. It wins 4 of the 4 coding tests both models report; Gemma 4 E4B wins none.
Which is cheaper?
Gemma 4 E4B costs $0.02 per million input tokens and $0.10 per million output tokens; Qwen3.5 9B costs $0.14 and $0.20. That makes Gemma 4 E4B about 64% cheaper for the same work.
How do you compare the two?
We use the 32 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 16 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.