Enter opens the highlighted result. Esc closes.

Claude Opus 5 vs Claude Sonnet 4.6

Updated October 21, 2026. 106 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace Claude Opus 5 Replace Claude Sonnet 4.6
Claude Opus 5Anthropic
Coding
81.8
Overall
50.8
Output $/M
$25.00
104
Claude Sonnet 4.6Anthropic
Coding
57.8
Overall
30.1
Output $/M
$15.00
2
vs boards won of 106 shared

Decision reading

Pick Claude Opus 5 for coding: it scores 81.8 on the Vibe Coding Index against 57.8 for Claude Sonnet 4.6.

Pick Claude Sonnet 4.6 when output tokens cost 1.7x less, or when what matters is TaxEval (Vals AI) or LiveBench data analysis, where it leads.

Model AClaude Opus 5Anthropic81.8/1002 independent benchmarks$1.33 per solved taskModel BClaude Sonnet 4.6Anthropic57.8/1002 independent benchmarks$1.68 per solved task
Shared results
106
Claude Opus 5 only
24
Claude Sonnet 4.6 only
11
Like-for-like categories
8 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
Claude Opus 5 81.8 from 2 benchmarks, Claude Sonnet 4.6 57.8 from 2 benchmarks.
Widest category
Math, 22.3 points apart on average over 4 shared benchmarks.
Widest single gap
ProofBench: Claude Opus 5 99.0% against Claude Sonnet 4.6 45.0%, 54.0 points apart on a 0 to 100 scale.
Output price
Claude Opus 5 $25.00 against Claude Sonnet 4.6 $15.00 per million output tokens, 1.7x apart.
Context window
Both take 1M tokens.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryClaude Opus 5BarsClaude Sonnet 4.6Ties
Coding180
Intelligence20
Agentic work170
Reasoning150
Math40
Knowledge91
Web design90
General301
All shared boards1042
CodingLike-for-likeFour of these boards feed the Vibe Coding Index180
Coding benchmarks for Claude Opus 5 and Claude Sonnet 4.6
BenchmarkClaude Opus 5Claude Sonnet 4.6Leader
LiveBench coding90.8%66.6%Claude Opus 5
LMArena WebDev16891523Claude Opus 5
AA Coding Index78.063.0Claude Opus 5
LiveCodeBench (Vals AI)89.0%82.1%Claude Opus 5
SWE-bench Verified (Vals AI)97.0%77.4%Claude Opus 5
Terminal-Bench 4 (Vals AI)53.5%Not measuredOnly Claude Opus 5 measured
Vibe Code Bench (Vals AI)88.4%55.8%Claude Opus 5
Code migration (Vals AI)57.5%39.9%Claude Opus 5
ProgramBench (Vals AI)3.0%0.5%Claude Opus 5
IOI (Vals AI)84.3%Not measuredOnly Claude Opus 5 measured
Terminal-Bench 2.1 (Vals AI)84.6%57.3%Claude Opus 5
Vibe Code Bench 1-100 (Vals AI)28.5%Not measuredOnly Claude Opus 5 measured
SkillsBench (Vals AI)60.4%49.0%Claude Opus 5
CyberBench (Vals AI)65.4%Not measuredOnly Claude Opus 5 measured
SciCode56.4%50.1%Claude Opus 5
Terminal-Bench Hard (Artificial Analysis run)Not measured53.0%Only Claude Sonnet 4.6 measured
Terminal-BenchNot measured53.4%Only Claude Sonnet 4.6 measured
FrontierCode53.4%24.3%Claude Opus 5
FrontierSWE52.0%Not measuredOnly Claude Opus 5 measured
WeirdML91.8%66.1%Claude Opus 5
ALE-Bench2164.6Not measuredOnly Claude Opus 5 measured
LMArena Text: coding15331503Claude Opus 5
LMArena Text: software and IT15271491Claude Opus 5
LMArena WebDev: React17111512Claude Opus 5
LMArena WebDev: HTML16801527Claude Opus 5
LMArena WebDev: from an image16621529Claude Opus 5
Neither model has a result on the other 12 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for Claude Opus 5 and Claude Sonnet 4.6
BenchmarkClaude Opus 5Claude Sonnet 4.6Leader
AA Intelligence Index50.830.1Claude Opus 5
Epoch Capabilities Index163.9152.8Claude Opus 5
Agentic workLike-for-likePublic boards, shown as published, kept out of the index170
Agentic work benchmarks for Claude Opus 5 and Claude Sonnet 4.6
BenchmarkClaude Opus 5Claude Sonnet 4.6Leader
AA Agentic Index56.531.8Claude Opus 5
Design Arena: agents building web apps12661190Claude Opus 5
Design Arena: agents building full-stack apps13061213Claude Opus 5
Design Arena: agents building mobile apps13481212Claude Opus 5
Design Arena: agents building Android apps12581202Claude Opus 5
Design Arena: agents building Godot games12791226Claude Opus 5
Design Arena: agents building games12611177Claude Opus 5
Design Arena: PowerPoint slides1265Not measuredOnly Claude Opus 5 measured
LiveBench research88.9%63.6%Claude Opus 5
LiveBench browser84.0%68.0%Claude Opus 5
LiveBench long agentic62.1%23.1%Claude Opus 5
SREBench (Vals AI)12.2%Not measuredOnly Claude Opus 5 measured
Finance Agent (Vals AI)58.6%51.0%Claude Opus 5
Tax Agent Bench (Vals AI)45.5%25.8%Claude Opus 5
Legal Research Bench (Vals AI)55.3%38.5%Claude Opus 5
Excel Modeling Benchmark (Vals AI)73.6%60.2%Claude Opus 5
BioMysteryBench (Vals AI)79.3%Not measuredOnly Claude Opus 5 measured
Terminal-Bench-Science (Vals AI)27.1%Not measuredOnly Claude Opus 5 measured
Vals Science Index47.9%Not measuredOnly Claude Opus 5 measured
Time Horizon Index (Vals AI)18.8%Not measuredOnly Claude Opus 5 measured
CUA-bench (Vals AI)9.0%Not measuredOnly Claude Opus 5 measured
τ²-Bench TelecomNot measured79.5%Only Claude Sonnet 4.6 measured
APEX-Agents65.8%43.0%Claude Opus 5
OSWorld 2.031.4%9.3%Claude Opus 5
BALROG63.4%Not measuredOnly Claude Opus 5 measured
DeepResearch BenchNot measured54.9%Only Claude Sonnet 4.6 measured
PostTrainBench35.0%Not measuredOnly Claude Opus 5 measured
OSWorldNot measured72.1%Only Claude Sonnet 4.6 measured
Vending-Bench 2$11,182$7,204Claude Opus 5
LMArena Search1242Not measuredOnly Claude Opus 5 measured
Neither model has a result on the other 9 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index150
Reasoning benchmarks for Claude Opus 5 and Claude Sonnet 4.6
BenchmarkClaude Opus 5Claude Sonnet 4.6Leader
GPQA Diamond93.4%85.6%Claude Opus 5
LiveBench reasoning91.2%86.4%Claude Opus 5
MysteryMechanism (Vals AI)37.4%Not measuredOnly Claude Opus 5 measured
MMMU Pro (Vals AI)89.9%83.6%Claude Opus 5
GPQA Diamond (Artificial Analysis run)93.7%87.5%Claude Opus 5
Humanity's Last Exam (Artificial Analysis run)54.9%33.6%Claude Opus 5
GPQA Diamond (Epoch AI run)93.9%87.4%Claude Opus 5
ARC-AGI97.5%86.5%Claude Opus 5
ARC-AGI-290.4%60.4%Claude Opus 5
SimpleBench80.6%Not measuredOnly Claude Opus 5 measured
Chess puzzles43.0%23.0%Claude Opus 5
Mystery game puzzles59.0%16.0%Claude Opus 5
EBR-bench45.7%Not measuredOnly Claude Opus 5 measured
LMArena Text: hard prompts15271485Claude Opus 5
LMArena Text: expert15541500Claude Opus 5
LMArena Text: hard prompts in English15231493Claude Opus 5
LMArena Vision: diagrams13481306Claude Opus 5
LMArena Vision: homework13381311Claude Opus 5
Neither model has a result on the other 3 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index40
Math benchmarks for Claude Opus 5 and Claude Sonnet 4.6
BenchmarkClaude Opus 5Claude Sonnet 4.6Leader
AIMENot measured92.3%Only Claude Sonnet 4.6 measured
OTIS Mock AIME98.9%85.8%Claude Opus 5
FrontierMath, tiers 1 to 385.6%Not measuredOnly Claude Opus 5 measured
FrontierMath, tier 473.2%Not measuredOnly Claude Opus 5 measured
ProofBench99.0%45.0%Claude Opus 5
LMArena Text: math15331462Claude Opus 5
LMArena Text: mathematical work15441483Claude Opus 5
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index91
Knowledge benchmarks for Claude Opus 5 and Claude Sonnet 4.6
BenchmarkClaude Opus 5Claude Sonnet 4.6Leader
MMLU Pro91.6%87.3%Claude Opus 5
LegalBench (Vals AI)87.0%82.1%Claude Opus 5
CaseLaw (Vals AI)Not measured64.0%Only Claude Sonnet 4.6 measured
CorpFin (Vals AI)73.2%65.3%Claude Opus 5
TaxEval (Vals AI)75.1%77.1%Claude Sonnet 4.6
MortgageTax (Vals AI)72.1%67.7%Claude Opus 5
MedQA (Vals AI)Not measured92.1%Only Claude Sonnet 4.6 measured
MedCode (Vals AI)63.6%Not measuredOnly Claude Opus 5 measured
SimpleQA Verified59.9%35.5%Claude Opus 5
LMArena Text: science15311474Claude Opus 5
LMArena Text: medicine and healthcare15251451Claude Opus 5
LMArena Text: legal and government15301470Claude Opus 5
LMArena Text: business and finance15011464Claude Opus 5
Neither model has a result on the other 4 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index90
Web design benchmarks for Claude Opus 5 and Claude Sonnet 4.6
BenchmarkClaude Opus 5Claude Sonnet 4.6Leader
Design Arena: websites13091285Claude Opus 5
Design Arena: UI components13481271Claude Opus 5
Design Arena: game dev13391259Claude Opus 5
Design Arena: data viz13291278Claude Opus 5
Design Arena: 3D13351234Claude Opus 5
Design Arena: all code categories13241276Claude Opus 5
Design Arena: SVG13101196Claude Opus 5
Design Arena: ASCII art13521235Claude Opus 5
LiveBench design83.0%71.8%Claude Opus 5
GeneralLike-for-likePublic boards, shown as published, kept out of the index301
General benchmarks for Claude Opus 5 and Claude Sonnet 4.6
BenchmarkClaude Opus 5Claude Sonnet 4.6Leader
LiveBench data analysis74.6%76.1%Claude Sonnet 4.6
Public Benefits Bench (Vals AI)76.9%62.5%Claude Opus 5
SAGE (Vals AI)49.4%46.6%Claude Opus 5
MedScribe (Vals AI)91.0%Not measuredOnly Claude Opus 5 measured
Vals Multimodal Index73.9%60.6%Claude Opus 5
IFBenchNot measured56.6%Only Claude Sonnet 4.6 measured
Long context reasoning (AA-LCR)82.0%80.0%Claude Opus 5
DTBench97.9%89.9%Claude Opus 5
LMCA64.5%46.5%Claude Opus 5
Furniture assembly60.8%Not measuredOnly Claude Opus 5 measured
ExploitBenchNot measured23.6%Only Claude Sonnet 4.6 measured
LMArena Text15091458Claude Opus 5
LMArena Text: instruction following15181476Claude Opus 5
LMArena Text: creative writing14931435Claude Opus 5
LMArena Vision13201283Claude Opus 5
LMArena Text: English15081472Claude Opus 5
LMArena Text: non-English15021440Claude Opus 5
LMArena Text: Chinese15731491Claude Opus 5
LMArena Text: German15271428Claude Opus 5
LMArena Text: Japanese15201420Claude Opus 5
LMArena Text: Korean15171410Claude Opus 5
LMArena Text: French15191465Claude Opus 5
LMArena Text: Spanish15231464Claude Opus 5
LMArena Text: Russian15071440Claude Opus 5
LMArena Text: Polish15221443Claude Opus 5
LMArena Text: multi-turn15001464Claude Opus 5
LMArena Text: longer queries15171479Claude Opus 5
LMArena Text: writing and language14991444Claude Opus 5
LMArena Text: entertainment and media14811431Claude Opus 5
LMArena Vision: English13151298Claude Opus 5
LMArena Vision: Chinese14001307Claude Opus 5
LMArena Vision: reading text13291296Claude Opus 5
LMArena Vision: humor12931248Claude Opus 5
LMArena Vision: recognizing thingsNot measured1195Only Claude Sonnet 4.6 measured
LMArena Vision: creative writing13131272Claude Opus 5
LMArena Document15161482Claude Opus 5
Neither model has a result on the other 10 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Claude Opus 5 and Claude Sonnet 4.6
BenchmarkClaude Opus 5Claude Sonnet 4.6Leader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

Claude Opus 5$0.018Fits in one request
Claude Sonnet 4.6$0.011Fits in one request

Claude Sonnet 4.6 costs less

Repository review

50K input + 3K output tokens

Claude Opus 5$0.325Fits in one request
Claude Sonnet 4.6$0.195Fits in one request

Claude Sonnet 4.6 costs less

Agent loop

200K input + 10K output tokens

Claude Opus 5$1.25Fits in one request
Claude Sonnet 4.6$0.750Fits in one request

Claude Sonnet 4.6 costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of Claude Opus 5 and Claude Sonnet 4.6
SpecificationAnthropic Claude Opus 5Anthropic Claude Sonnet 4.6
MakerAnthropicAnthropic
WeightsNo public weights listedNo public weights listed
ReleasedJuly 24, 2026February 17, 2026
Context window1M1M
Input $/Mtok$5.00$3.00
Output $/Mtok$25.00$15.00
Coding index81.857.8
Independent measurements behind it22
Cost per solved task$1.33$1.68
SourceOpenRouter listingOpenRouter listing

Questions

Is Claude Opus 5 better than Claude Sonnet 4.6 for coding?

Pick Claude Opus 5 for coding: it scores 81.8 on the Vibe Coding Index against 57.8 for Claude Sonnet 4.6. Pick Claude Sonnet 4.6 when output tokens cost 1.7x less, or when what matters is TaxEval (Vals AI) or LiveBench data analysis, where it leads.

Which is cheaper, Claude Opus 5 or Claude Sonnet 4.6?

Claude Opus 5 costs $5.00 per million input tokens and $25.00 per million output tokens; Claude Sonnet 4.6 costs $3.00 and $15.00. Per solved SWE-bench Verified task, Claude Opus 5 costs $1.33 and Claude Sonnet 4.6 $1.68.

Which has the bigger context window, Claude Opus 5 or Claude Sonnet 4.6?

Neither: both take 1M tokens.

How many benchmarks is the Claude Opus 5 vs Claude Sonnet 4.6 comparison based on?

106 benchmarks have a result for both models as of October 21, 2026. Claude Opus 5's coding index rests on 2 independent benchmarks and Claude Sonnet 4.6's on 2. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/claude-opus-5-vs-claude-sonnet-4-6" title="Claude Opus 5 vs Claude Sonnet 4.6, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/claude-opus-5-vs-claude-sonnet-4-6">LLMBenchmark.ai, Claude Opus 5 vs Claude Sonnet 4.6</a></p>