Claude Sonnet 4.5 vs Gemma 4 26B A4B
Wins 4 of 8 areas
Coding · Reasoning · Facts · Long documents
Wins 2 of 8 areas
Math · Following instructions
Claude Sonnet 4.5 wins more areas, narrowly.Gemma 4 26B A4B is cheaper and better at math.
Scores updated · 49 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 software71Claude Sonnet 4.57 of 8 tests
- ReasoningHard problems that need careful thinking21Claude Sonnet 4.52 of 3 tests
- FactsGetting facts right instead of making them up21Claude Sonnet 4.52 of 3 tests
- Long documentsFinding answers in very long texts10Claude Sonnet 4.51 of 1 test
- MathCompetition and research-level math01Gemma 4 26B A4B1 of 1 test
- Following instructionsDoing exactly what it is asked01Gemma 4 26B A4B1 of 1 test
- AgentsCarrying out multi-step tasks on its own22Even2 each
- Images and chartsUnderstanding pictures, charts and video11Even1 each · 2 ties
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 26B A4B costs 97% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Claude Sonnet 4.5 pulls ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+59.8points ahead
- Fixes real GitHub issues in many programming languagesSWE-bench Multilingual+49.7points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+37.3points ahead
Where Gemma 4 26B A4B pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+15.1points ahead
- Recent programming contest problemsLiveCodeBench v6+13.1points ahead
- Olympiad math problems with short answersIMOAnswerBench+8.4points ahead
Every test, side by side
All 49 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Sonnet 4.5
- SWE-bench VerifiedClaude Sonnet 4.5 by 59.877.217.4+59.8
- SWE-bench MultilingualClaude Sonnet 4.5 by 49.76717.3+49.7
- SWE-bench ProClaude Sonnet 4.5 by 29.843.613.8+29.8
- Terminal-Bench HardClaude Sonnet 4.5 by 2235.613.6+22
- Terminal-Bench 2.1Claude Sonnet 4.5 by 16.855.839+16.8
- LiveCodeBench v6Gemma 4 26B A4B by 13.16477.1+13.1
- SciCodeClaude Sonnet 4.5 by 5.745.740+5.7
- LMArena · WebDevClaude Sonnet 4.5 by 35 rating points13941359+35 rating
AgentsEven
- GDPValClaude Sonnet 4.5 by 17.120.53.4+17.1
- τ-Bench V3 · BankingClaude Sonnet 4.5 by 12.524.512+12.5
- MCP AtlasGemma 4 26B A4B by 6.243.850+6.2
- BrowseCompGemma 4 26B A4B by 2.224.126.3+2.2
ReasoningClaude Sonnet 4.5
- GPQA DiamondClaude Sonnet 4.5 by 4.283.479.2+4.2
- Humanity's Last ExamGemma 4 26B A4B by 1.517.819.3+1.5
- CritPtClaude Sonnet 4.5 by 1.11.10+1.1
FactsClaude Sonnet 4.5
- AA-Omniscience · Non-hallucinationClaude Sonnet 4.5 by 37.350.913.6+37.3
- AA-Omniscience · AccuracyClaude Sonnet 4.5 by 13.832.919.1+13.8
- Vectara HHEM hallucination ratelower is betterGemma 4 26B A4B by 6.8125.2+6.8
Images and chartsEven
- CharXiv (RQ)Gemma 4 26B A4B by 1.867.269+1.8
- MMMUClaude Sonnet 4.5 by 1.279.678.4+1.2
- MMMU-Protie68.769.2tie
- MathVistatie79.879.4tie
Long documentsClaude Sonnet 4.5
- AA-LCRClaude Sonnet 4.5 by 6.672.365.7+6.6
Following instructionsGemma 4 26B A4B
- IFBenchGemma 4 26B A4B by 15.157.372.4+15.1
Other results24 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.
- AA-OmniscienceClaude Sonnet 4.5 by 50.7-0.1-50.8+50.7
- τ²-Bench Telecom (AA run)Claude Sonnet 4.5 by 34.578.143.6+34.5
- τ²-BenchClaude Sonnet 4.5 by 29.89868.2+29.8
- HMMT Feb. 2025Gemma 4 26B A4B by 24.267.591.7+24.2
- Terminal-Bench 2.0Claude Sonnet 4.5 by 15.85034.2+15.8
- HLE (with tools)Claude Sonnet 4.5 by 14.83217.2+14.8
- Artificial Analysis Coding IndexClaude Sonnet 4.5 by 12.852.139.3+12.8
- OmniDocBench 1.5Claude Sonnet 4.5 by 11.485.874.4+11.4
- vectara_factual_consistencyGemma 4 26B A4B by 6.88894.8+6.8
- AA Agentic IndexClaude Sonnet 4.5 by 6.517.511+6.5
- CC-OCRGemma 4 26B A4B by 6.468.174.5+6.4
- HallusionBenchGemma 4 26B A4B by 6.259.966.1+6.2
- vectara_avg_summary_lengthClaude Sonnet 4.5 by 60.7 rating points127.867.1+60.7 rating
- HMMT Nov. 2025Gemma 4 26B A4B by 5.881.787.5+5.8
- SimpleVQAClaude Sonnet 4.5 by 5.457.652.2+5.4
- MMLU-ProClaude Sonnet 4.5 by 4.987.582.6+4.9
- vectara_answer_rateGemma 4 26B A4B by 4.295.699.8+4.2
- AA IntelligenceClaude Sonnet 4.5 by 420.716.7+4
- MMLU-ReduxClaude Sonnet 4.5 by 2.995.692.7+2.9
- MMMLUClaude Sonnet 4.5 by 2.889.186.3+2.8
- RealWorldQAGemma 4 26B A4B by 1.970.372.2+1.9
- AI2DGemma 4 26B A4B by 1.38788.3+1.3
- GPQA (unspecified)Claude Sonnet 4.5 by 1.183.482.3+1.1
- MMBench EN-DEV-v1.1tie88.389tie
Questions people ask
Which is better, Claude Sonnet 4.5 or Gemma 4 26B A4B?
Claude Sonnet 4.5 wins four of the eight areas we test: coding, reasoning, facts and long documents. Gemma 4 26B A4B wins math and following instructions, and costs 97% less. They are level on agents and images and charts.
Which is better for coding?
Claude Sonnet 4.5. It wins 7 of the 8 coding tests both models report; Gemma 4 26B A4B wins 1.
Which is cheaper?
Claude Sonnet 4.5 costs $3.00 per million input tokens and $15.00 per million output tokens; Gemma 4 26B A4B costs $0.13 and $0.40. That makes Gemma 4 26B A4B about 97% cheaper for the same work.
How do you compare the two?
We use the 49 benchmark tests both models have published scores on. The verdict counts the 25 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 24 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.