Gemma 4 26B A4B vs Qwen3.5 122B A10B
Wins 0 of 7 areas
—
Wins 6 of 7 areas
Coding · Agents · Reasoning · Images and charts · Long documents · Following instructions
Qwen3.5 122B A10B is the stronger all-rounder.Gemma 4 26B A4B is cheaper.
Scores updated · 53 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 software06Qwen3.5 122B A10B6 of 8 tests · 2 ties
- Images and chartsUnderstanding pictures, charts and video14Qwen3.5 122B A10B4 of 5 tests
- AgentsCarrying out multi-step tasks on its own03Qwen3.5 122B A10B3 of 3 tests
- ReasoningHard problems that need careful thinking02Qwen3.5 122B A10B2 of 3 tests · 1 tie
- Long documentsFinding answers in very long texts01Qwen3.5 122B A10B1 of 1 test
- Following instructionsDoing exactly what it is asked01Qwen3.5 122B A10B1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each · 1 tie
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 26B A4B costs 85% 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 122B A10B pulls ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+54.6points ahead
- Fixes real GitHub issues in many programming languagesSWE-bench Multilingual+44.7points ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+37.5points ahead
Where Gemma 4 26B A4B pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 53 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingQwen3.5 122B A10B
- SWE-bench VerifiedQwen3.5 122B A10B by 54.617.472+54.6
- SWE-bench MultilingualQwen3.5 122B A10B by 44.717.362+44.7
- SWE-bench ProQwen3.5 122B A10B by 31.113.844.9+31.1
- Terminal-Bench HardQwen3.5 122B A10B by 17.513.631.1+17.5
- Terminal-Bench 2.1Qwen3.5 122B A10B by 8.63947.6+8.6
- LiveCodeBench v6Qwen3.5 122B A10B by 1.877.178.9+1.8
- SciCodetie4039.7tie
- LMArena · WebDevtie13591360tie
AgentsQwen3.5 122B A10B
- BrowseCompQwen3.5 122B A10B by 37.526.363.8+37.5
- GDPValQwen3.5 122B A10B by 12.43.315.7+12.4
- τ-Bench V3 · BankingQwen3.5 122B A10B by 3.31215.3+3.3
ReasoningQwen3.5 122B A10B
- GPQA DiamondQwen3.5 122B A10B by 6.579.285.7+6.5
- Humanity's Last ExamQwen3.5 122B A10B by 5.919.325.2+5.9
- CritPttie00.6tie
FactsEven
- Vectara HHEM hallucination ratelower is betterGemma 4 26B A4B by 65.211.2+6
- AA-Omniscience · AccuracyQwen3.5 122B A10B by 5.319.124.4+5.3
- AA-Omniscience · Non-hallucinationtie13.612.9tie
Images and chartsQwen3.5 122B A10B
- CharXiv (RQ)Qwen3.5 122B A10B by 8.26977.2+8.2
- MathVistaQwen3.5 122B A10B by 879.487.4+8
- MMMU-ProQwen3.5 122B A10B by 5.869.275+5.8
- MMMUQwen3.5 122B A10B by 5.578.483.9+5.5
- LMArena · VisionGemma 4 26B A4B by 15 rating points12601245+15 rating
Long documentsQwen3.5 122B A10B
- AA-LCRQwen3.5 122B A10B by 10.665.776.3+10.6
Following instructionsQwen3.5 122B A10B
- IFBenchQwen3.5 122B A10B by 3.372.475.7+3.3
Other results29 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 122B A10B by 5043.693.6+50
- WideSearchQwen3.5 122B A10B by 22.238.360.5+22.2
- vectara_avg_summary_lengthQwen3.5 122B A10B by 19.367.186.4+19.3
- OmniDocBench 1.5Qwen3.5 122B A10B by 15.474.489.8+15.4
- Terminal-Bench 2.0Qwen3.5 122B A10B by 15.234.249.4+15.2
- BFCLv4Qwen3.5 122B A10B by 15.157.172.2+15.1
- RealWorldQAQwen3.5 122B A10B by 12.972.285.1+12.9
- τ³-Bench BankingQwen3.5 122B A10B by 9.74.514.2+9.7
- SimpleVQAQwen3.5 122B A10B by 9.552.261.7+9.5
- C-EvalQwen3.5 122B A10B by 9.482.591.9+9.4
- AA-OmniscienceQwen3.5 122B A10B by 9.3-50.8-41.5+9.3
- DeepPlanningQwen3.5 122B A10B by 7.916.224.1+7.9
- CC-OCRQwen3.5 122B A10B by 7.374.581.8+7.3
- Artificial Analysis Coding IndexQwen3.5 122B A10B by 6.439.345.7+6.4
- t2-benchGemma 4 26B A4B by 685.579.5+6
- vectara_factual_consistencyGemma 4 26B A4B by 694.888.8+6
- SuperGPQAQwen3.5 122B A10B by 5.761.467.1+5.7
- GDPval-AA v2Qwen3.5 122B A10B by 54 rating points807861+54 rating
- AI2DQwen3.5 122B A10B by 588.393.3+5
- MMLU-ProQwen3.5 122B A10B by 4.182.686.7+4.1
- MATH-VisionQwen3.5 122B A10B by 3.882.486.2+3.8
- VITA-BenchGemma 4 26B A4B by 3.336.933.6+3.3
- HallusionBenchQwen3.5 122B A10B by 1.566.167.6+1.5
- AA Agentic IndexGemma 4 26B A4B by 1.4119.6+1.4
- IFEvalQwen3.5 122B A10B by 1.49293.4+1.4
- MMLU-ReduxQwen3.5 122B A10B by 1.392.794+1.3
- AA IntelligenceGemma 4 26B A4B by 1.116.715.6+1.1
- MMMLUtie86.386.7tie
- vectara_answer_ratetie99.899.8tie
Questions people ask
Which is better, Gemma 4 26B A4B or Qwen3.5 122B A10B?
Qwen3.5 122B A10B wins six of the seven areas where both have results: coding, agents, reasoning, images and charts, long documents and following instructions. Gemma 4 26B A4B wins none, but costs 85% less. They are level on facts.
Which is better for coding?
Qwen3.5 122B A10B. It wins 6 of the 8 coding tests both models report; Gemma 4 26B A4B wins none, and 2 are ties.
Which is cheaper?
Gemma 4 26B A4B costs $0.13 per million input tokens and $0.40 per million output tokens; Qwen3.5 122B A10B costs $0.40 and $3.20. That makes Gemma 4 26B A4B about 85% cheaper for the same work.
How do you compare the two?
We use the 53 benchmark tests both models have published scores on. The verdict counts the 24 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 29 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.