Claude Sonnet 5 vs Claude Opus 5
Wins 1 of 8 areas
Long documents
Wins 6 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math
Claude Opus 5 is the stronger all-rounder.Claude Sonnet 5 is cheaper and better at long documents.
Scores updated · 61 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 software07Claude Opus 57 of 8 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own05Claude Opus 55 of 5 tests
- ReasoningHard problems that need careful thinking05Claude Opus 55 of 5 tests
- MathCompetition and research-level math03Claude Opus 53 of 3 tests
- FactsGetting facts right instead of making them up12Claude Opus 52 of 3 tests
- Images and chartsUnderstanding pictures, charts and video12Claude Opus 52 of 3 tests
- Long documentsFinding answers in very long texts10Claude Sonnet 51 of 1 test
- Following instructionsDoing exactly what it is asked00Even0 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.
Claude Sonnet 5 costs 60% 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 Opus 5 pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+43.9points ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+34.9points ahead
- Short factual questions, answered correctlySimpleQA Verified+26.2points ahead
Where Claude Sonnet 5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+21.4points ahead
- Reasoning about charts from research papersCharXiv (RQ)+4.6points ahead
- Reasons across sets of long documentsAA-LCR+2.7points ahead
Every test, side by side
All 61 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 5
- SWE-bench ProClaude Opus 5 by 1663.279.2+16
- LMArena · WebDevClaude Opus 5 by 151 rating points15411692+151 rating
- SWE-bench MultilingualClaude Opus 5 by 11.278.389.5+11.2
- SWE-bench VerifiedClaude Opus 5 by 10.885.296+10.8
- Terminal-Bench 2.1Claude Opus 5 by 8.680.589.1+8.6
- LiveBench · Agentic CodingClaude Opus 5 by 5.859.465.2+5.8
- SciCodeClaude Opus 5 by 2.154.356.4+2.1
- LiveBench · Codingtie80.781.5tie
AgentsClaude Opus 5
- Terminal-Bench 4.0Claude Opus 5 by 34.914.149+34.9
- GDPValClaude Opus 5 by 12.948.361.2+12.9
- BrowseCompClaude Opus 5 by 6.184.790.8+6.1
- τ-Bench V3 · BankingClaude Opus 5 by 4.837.342.1+4.8
- OSWorld-VerifiedClaude Opus 5 by 2.281.283.4+2.2
ReasoningClaude Opus 5
- SimpleBenchClaude Opus 5 by 2060.680.6+20
- Humanity's Last ExamClaude Opus 5 by 13.641.354.9+13.6
- CritPtClaude Opus 5 by 12.216.929.1+12.2
- LiveBench · ReasoningClaude Opus 5 by 2.588.791.2+2.5
- GPQA DiamondClaude Opus 5 by 2.191.193.2+2.1
FactsClaude Opus 5
- SimpleQA VerifiedClaude Opus 5 by 26.233.759.9+26.2
- AA-Omniscience · Non-hallucinationClaude Sonnet 5 by 21.460.639.2+21.4
- AA-Omniscience · AccuracyClaude Opus 5 by 20.84060.9+20.8
Images and chartsClaude Opus 5
- MMMU-ProClaude Opus 5 by 7.477.384.7+7.4
- CharXiv (RQ)Claude Sonnet 5 by 4.688.383.7+4.6
- LMArena · VisionClaude Opus 5 by 46 rating points12751321+46 rating
MathClaude Opus 5
- FrontierMath Tier 4Claude Opus 5 by 43.929.373.2+43.9
- FrontierMath Tiers 1-3 (v2)Claude Opus 5 by 2065.685.6+20
- LiveBench · MathematicsClaude Opus 5 by 2.892.995.7+2.8
Following instructionsEven
- LiveBench · Instruction Followingtie63.963.8tie
Other results32 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.
- Terminal-Bench 4.0Claude Opus 5 by 39.412.451.8+39.4
- OSWorld 2.0 (partial)Claude Opus 5 by 32.842.675.4+32.8
- SWE-bench MultimodalClaude Opus 5 by 31.328.159.4+31.3
- GDPval-AA 2.1Claude Opus 5 by 259 rating points14491708+259 rating
- GDPval-AA v2 EloClaude Opus 5 by 240 rating points15841824+240 rating
- HLE-VerifiedClaude Opus 5 by 23.43154.4+23.4
- AA-OmniscienceClaude Opus 5 by 20.616.437.1+20.6
- GDPval-AA v2Claude Opus 5 by 206 rating points16181824+206 rating
- OSWorld 2.1 partialClaude Opus 5 by 175774+17
- DeepSWE 1.1Claude Opus 5 by 14.85468.8+14.8
- livebench_languageClaude Opus 5 by 13.77588.7+13.7
- AA IntelligenceClaude Opus 5 by 12.638.250.8+12.6
- AA Agentic IndexClaude Opus 5 by 11.944.356.2+11.9
- ProteinGym HardClaude Opus 5 by 11.136.647.7+11.1
- HealthBench Professional (raw score)Claude Opus 5 by 1162.473.4+11
- HealthBench Professional (raw)Claude Opus 5 by 1162.473.4+11
- GDP.PDF (All pass rate)Claude Opus 5 by 92837+9
- Program BenchClaude Opus 5 by 8.177.385.4+8.1
- HealthBench (raw score)Claude Opus 5 by 7.959.267.1+7.9
- HealthBench (raw)Claude Opus 5 by 7.959.267.1+7.9
- OfficeQA ProClaude Opus 5 by 7.559.466.9+7.5
- LVBenchClaude Opus 5 by 6.968.575.4+6.9
- Artificial Analysis Coding IndexClaude Opus 5 by 6.571.578+6.5
- HLE (with tools)Claude Opus 5 by 6.257.463.6+6.2
- Legal Agent BenchmarkClaude Opus 5 by 5.95.811.7+5.9
- Toolathlon VerifiedClaude Opus 5 by 5.974.780.6+5.9
- GDP (Surge AI)Claude Opus 5 by 3.981.685.5+3.9
- GMMLUClaude Opus 5 by 3.389.292.5+3.3
- OfficeQAClaude Opus 5 by 375.178.1+3
- livebench_data_analysisClaude Opus 5 by 2.971.774.5+2.9
- MILUClaude Opus 5 by 2.889.392.1+2.8
- HealthBench ProfessionalClaude Opus 5 by 257.859.8+2
Questions people ask
Which is better, Claude Sonnet 5 or Claude Opus 5?
Claude Opus 5 wins six of the eight areas we test: coding, agents, reasoning, facts, images and charts and math. Claude Sonnet 5 wins long documents, and costs 60% less. They are level on following instructions.
Which is better for coding?
Claude Opus 5. It wins 7 of the 8 coding tests both models report; Claude Sonnet 5 wins none, and 1 is a tie.
Which is cheaper?
Claude Sonnet 5 costs $2.00 per million input tokens and $10.00 per million output tokens; Claude Opus 5 costs $5.00 and $25.00. That makes Claude Sonnet 5 about 60% cheaper for the same work.
How do you compare the two?
We use the 61 benchmark tests both models have published scores on. The verdict counts the 29 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 32 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.