Claude Sonnet 5 vs Gemini 3.7 Flash
Wins 1 of 8 areas
Agents
Wins 7 of 8 areas
Coding · Reasoning · Facts · Images and charts · Math · Long documents · Following instructions
Gemini 3.7 Flash is the stronger all-rounder.Claude Sonnet 5 is better at agents.
Scores updated · 43 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.
- Images and chartsUnderstanding pictures, charts and video02Gemini 3.7 Flash2 of 3 tests · 1 tie
- MathCompetition and research-level math02Gemini 3.7 Flash2 of 3 tests · 1 tie
- CodingWriting and fixing software23Gemini 3.7 Flash3 of 5 tests
- ReasoningHard problems that need careful thinking12Gemini 3.7 Flash2 of 4 tests · 1 tie
- FactsGetting facts right instead of making them up12Gemini 3.7 Flash2 of 3 tests
- Long documentsFinding answers in very long texts01Gemini 3.7 Flash1 of 2 tests · 1 tie
- Following instructionsDoing exactly what it is asked01Gemini 3.7 Flash1 of 1 test
- AgentsCarrying out multi-step tasks on its own20Claude Sonnet 52 of 3 tests · 1 tie
Following instructions rests on a single test.
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.
Gemini 3.7 Flash costs 63% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Gemini 3.7 Flash pulls ahead
- Short factual questions, answered correctlySimpleQA Verified+35.5points ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+16points ahead
- Finds one of eight look-alike replies in a long chatMRCR v2 8 needle 128k (average)+15.5points ahead
Where Claude Sonnet 5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+25.1points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+4.5points ahead
- Real work tasks from 44 professionsGDPVal+3.7points ahead
Every test, side by side
All 43 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 3.7 Flash
- LMArena · WebDevGemini 3.7 Flash by 53 rating points15391592+53 rating
- Terminal-Bench 2.1Gemini 3.7 Flash by 5.380.585.8+5.3
- SciCodeGemini 3.7 Flash by 2.954.357.2+2.9
- LiveBench · CodingClaude Sonnet 5 by 1.880.778.9+1.8
- LiveBench · Agentic CodingClaude Sonnet 5 by 1.159.458.3+1.1
AgentsClaude Sonnet 5
- τ-Bench V3 · BankingClaude Sonnet 5 by 4.537.332.8+4.5
- GDPValClaude Sonnet 5 by 3.748.344.6+3.7
- Terminal-Bench 4.0tie14.113.6tie
ReasoningGemini 3.7 Flash
- Humanity's Last ExamGemini 3.7 Flash by 6.641.347.9+6.6
- GPQA DiamondGemini 3.7 Flash by 3.491.194.5+3.4
- CritPtClaude Sonnet 5 by 2.616.914.3+2.6
- LiveBench · Reasoningtie88.787.8tie
FactsGemini 3.7 Flash
- SimpleQA VerifiedGemini 3.7 Flash by 35.533.769.2+35.5
- AA-Omniscience · Non-hallucinationClaude Sonnet 5 by 25.160.635.5+25.1
- AA-Omniscience · AccuracyGemini 3.7 Flash by 15.24055.3+15.2
Images and chartsGemini 3.7 Flash
- MMMU-ProGemini 3.7 Flash by 8.277.385.5+8.2
- LMArena · VisionGemini 3.7 Flash by 40 rating points12751315+40 rating
- CharXiv (RQ)tie88.388.7tie
MathGemini 3.7 Flash
- FrontierMath Tier 4Gemini 3.7 Flash by 7.329.336.6+7.3
- FrontierMath Tiers 1-3 (v2)Gemini 3.7 Flash by 665.671.6+6
- LiveBench · Mathematicstie92.993.5tie
Long documentsGemini 3.7 Flash
- MRCR v2 8 needle 128k (average)Gemini 3.7 Flash by 15.581.597+15.5
- AA-LCRtie8281.7tie
Following instructionsGemini 3.7 Flash
- LiveBench · Instruction FollowingGemini 3.7 Flash by 1663.979.9+16
Other results19 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.
- GDP (Surge AI)Claude Sonnet 5 by 47.681.634+47.6
- HLE-VerifiedGemini 3.7 Flash by 22.63153.6+22.6
- LVBenchGemini 3.7 Flash by 16.968.585.4+16.9
- DeepSWE 1.1Gemini 3.7 Flash by 11.35465.3+11.3
- livebench_languageGemini 3.7 Flash by 10.57585.5+10.5
- GDPval-AA v2 EloClaude Sonnet 5 by 102 rating points15841482+102 rating
- AA-OmniscienceGemini 3.7 Flash by 1016.426.5+10
- GDPval-AA v2Claude Sonnet 5 by 93 rating points16181525+93 rating
- OSWorld 2.0 (partial)Gemini 3.7 Flash by 842.650.6+8
- AA Agentic IndexClaude Sonnet 5 by 7.944.336.4+7.9
- Agent's Last Exam (Pass rate)Claude Sonnet 5 by 733.326.3+7
- GDP.PDF (All pass rate)Gemini 3.7 Flash by 62834+6
- Code Arena (Elo)Gemini 3.7 Flash by 47 rating points15411588+47 rating
- Artificial Analysis Coding IndexGemini 3.7 Flash by 4.671.576.1+4.6
- livebench_data_analysisClaude Sonnet 5 by 3.771.768+3.7
- Terminal-Bench 4.0Claude Sonnet 5 by 1.212.411.2+1.2
- AA Intelligencetie38.239.1tie
- CharXiv Reasoning (With tools)tie88.388.7tie
- Terminal-bench 3.0tie14.614.9tie
Questions people ask
Which is better, Claude Sonnet 5 or Gemini 3.7 Flash?
Gemini 3.7 Flash wins seven of the eight areas we test: coding, reasoning, facts, images and charts, math, long documents and following instructions. Claude Sonnet 5 wins agents.
Which is better for coding?
Gemini 3.7 Flash. It wins 3 of the 5 coding tests both models report; Claude Sonnet 5 wins 2.
Which is cheaper?
Claude Sonnet 5 costs $2.00 per million input tokens and $10.00 per million output tokens; Gemini 3.7 Flash costs $0.75 and $3.75. That makes Gemini 3.7 Flash about 63% cheaper for the same work.
How do you compare the two?
We use the 43 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 19 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.