Enter opens the highlighted result. Esc closes.

Claude Sonnet 4.5 vs Claude Sonnet 5

Claude Sonnet 5 leads Claude Sonnet 4.5 on coding, 57.7 to 39.5 on the Vibe Coding Index, and Claude Sonnet 5 costs 1.5x less per output token ($10.00 vs $15.00 per million). Updated October 22, 2026; 74 benchmarks have a result for both.

Replace Claude Sonnet 4.5 Replace Claude Sonnet 5
Claude Sonnet 4.5Anthropic
Coding
39.5
Overall
49.5
Output $/M
$15.00
15
Claude Sonnet 5Anthropic
Coding
57.7
Overall
63.2
Output $/M
$10.00
59
vs boards won of 74 shared

Claude Sonnet 4.5 or Claude Sonnet 5: which to pick

Pick Claude Sonnet 5 for coding: it scores 57.7 on the Vibe Coding Index against 39.5 for Claude Sonnet 4.5.

Pick Claude Sonnet 4.5 for a lower cost per solved task ($1.16 against $1.87) and work measured by LMArena Text: coding or LMArena Text: hard prompts, where it leads.

Model AClaude Sonnet 4.5Anthropic39.5/1003 independent benchmarks$1.16 per solved taskModel BClaude Sonnet 5Anthropic57.7/1006 independent benchmarks$1.87 per solved task
Shared results
74
Claude Sonnet 4.5 only
21
Claude Sonnet 5 only
47
Like-for-like categories
8 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

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 Sonnet 4.5 39.5 from 3 benchmarks, Claude Sonnet 5 57.7 from 6 benchmarks.
Widest category
Math, 25.0 points apart on average over 6 shared benchmarks.
Widest single gap
Vibe Code Bench (Vals AI): Claude Sonnet 4.5 22.6% against Claude Sonnet 5 81.3%, 58.7 points apart on a 0 to 100 scale.
Output price
Claude Sonnet 4.5 $15.00 against Claude Sonnet 5 $10.00 per million output tokens, 1.5x 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 Sonnet 4.5BarsClaude Sonnet 5Ties
Coding110
Intelligence02
Agentic work05
Reasoning27
Math06
Knowledge38
Web design08
General913
All shared boards1559
CodingLike-for-likeSix of these boards feed the Vibe Coding Index110
Coding benchmarks for Claude Sonnet 4.5 and Claude Sonnet 5
BenchmarkClaude Sonnet 4.5Claude Sonnet 5Leader
SWE-bench Verified74.8%Not measuredOnly Claude Sonnet 4.5 measured
SWE-bench Multilingual67.0%Not measuredOnly Claude Sonnet 4.5 measured
LiveBench codingNot measured74.7%Only Claude Sonnet 5 measured
LMArena WebDev13941540Claude Sonnet 5
AA Coding Index52.171.5Claude Sonnet 5
LiveCodeBench (Vals AI)73.0%82.4%Claude Sonnet 5
SWE-bench Verified (Vals AI)70.0%79.6%Claude Sonnet 5
Terminal-Bench 4 (Vals AI)Not measured10.6%Only Claude Sonnet 5 measured
Vibe Code Bench (Vals AI)22.6%81.3%Claude Sonnet 5
Code migration (Vals AI)Not measured44.4%Only Claude Sonnet 5 measured
ProgramBench (Vals AI)Not measured0.0%Only Claude Sonnet 5 measured
IOI (Vals AI)Not measured45.0%Only Claude Sonnet 5 measured
Terminal-Bench 2.1 (Vals AI)Not measured74.5%Only Claude Sonnet 5 measured
Vibe Code Bench 1-100 (Vals AI)Not measured13.8%Only Claude Sonnet 5 measured
SkillsBench (Vals AI)Not measured46.5%Only Claude Sonnet 5 measured
CyberBench (Vals AI)Not measured70.3%Only Claude Sonnet 5 measured
SciCodeNot measured54.3%Only Claude Sonnet 5 measured
Terminal-Bench46.5%Not measuredOnly Claude Sonnet 4.5 measured
FrontierCodeNot measured42.7%Only Claude Sonnet 5 measured
GSO-Bench14.7%37.3%Claude Sonnet 5
WeirdML47.7%68.8%Claude Sonnet 5
Cybench60.0%Not measuredOnly Claude Sonnet 4.5 measured
ALE-BenchNot measured1463.1Only Claude Sonnet 5 measured
LMArena Text: coding14881483Claude Sonnet 4.5
LMArena Text: software and IT14731474Claude Sonnet 5
LMArena WebDev: React13861541Claude Sonnet 5
LMArena WebDev: HTML13991527Claude Sonnet 5
LMArena WebDev: from an imageNot measured1540Only Claude Sonnet 5 measured
SciCodeNot measured10.7%Only Claude Sonnet 5 measured
Neither model has a result on the other 10 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for Claude Sonnet 4.5 and Claude Sonnet 5
BenchmarkClaude Sonnet 4.5Claude Sonnet 5Leader
AA Intelligence Index20.738.2Claude Sonnet 5
Epoch Capabilities Index146.7156.9Claude Sonnet 5
Agentic workLike-for-likePublic boards, shown as published, kept out of the index05
Agentic work benchmarks for Claude Sonnet 4.5 and Claude Sonnet 5
BenchmarkClaude Sonnet 4.5Claude Sonnet 5Leader
AA Agentic Index15.843.6Claude Sonnet 5
Design Arena: agents building web apps10601245Claude Sonnet 5
Design Arena: agents building full-stack apps10501249Claude Sonnet 5
Design Arena: agents building mobile apps11371215Claude Sonnet 5
Design Arena: agents building Android appsNot measured1227Only Claude Sonnet 5 measured
Design Arena: agents building Godot gamesNot measured1225Only Claude Sonnet 5 measured
Design Arena: agents building gamesNot measured1227Only Claude Sonnet 5 measured
Design Arena: HTML slidesNot measured1168Only Claude Sonnet 5 measured
Design Arena: PowerPoint slidesNot measured1191Only Claude Sonnet 5 measured
Design Arena: spreadsheetsNot measured1035Only Claude Sonnet 5 measured
LiveBench researchNot measured82.8%Only Claude Sonnet 5 measured
LiveBench browserNot measured72.0%Only Claude Sonnet 5 measured
LiveBench long agenticNot measured39.0%Only Claude Sonnet 5 measured
SREBench (Vals AI)Not measured0.4%Only Claude Sonnet 5 measured
Finance Agent (Vals AI)Not measured53.9%Only Claude Sonnet 5 measured
Tax Agent Bench (Vals AI)Not measured29.4%Only Claude Sonnet 5 measured
Legal Research Bench (Vals AI)Not measured41.8%Only Claude Sonnet 5 measured
Excel Modeling Benchmark (Vals AI)Not measured66.3%Only Claude Sonnet 5 measured
Terminal-Bench-Science (Vals AI)Not measured5.7%Only Claude Sonnet 5 measured
APEX-AgentsNot measured54.5%Only Claude Sonnet 5 measured
METR time horizons67.4%Not measuredOnly Claude Sonnet 4.5 measured
Remote Labor Index2.1%Not measuredOnly Claude Sonnet 4.5 measured
GDPval42.5%Not measuredOnly Claude Sonnet 4.5 measured
DeepResearch Bench52.6%Not measuredOnly Claude Sonnet 4.5 measured
OSWorld62.9%Not measuredOnly Claude Sonnet 4.5 measured
Vending-Bench 2$3,839$6,378Claude Sonnet 5
Berkeley Function Calling Leaderboard (BFCL)73.2%Not measuredOnly Claude Sonnet 4.5 measured
Neither model has a result on the other 14 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index27
Reasoning benchmarks for Claude Sonnet 4.5 and Claude Sonnet 5
BenchmarkClaude Sonnet 4.5Claude Sonnet 5Leader
GPQA Diamond81.6%88.9%Claude Sonnet 5
LiveBench reasoningNot measured88.7%Only Claude Sonnet 5 measured
MMMU Pro (Vals AI)79.3%83.0%Claude Sonnet 5
GPQA Diamond (Artificial Analysis run)Not measured91.1%Only Claude Sonnet 5 measured
Humanity's Last Exam (Artificial Analysis run)Not measured41.3%Only Claude Sonnet 5 measured
GPQA Diamond (Epoch AI run)82.3%90.5%Claude Sonnet 5
ARC-AGI63.7%Not measuredOnly Claude Sonnet 4.5 measured
ARC-AGI-213.6%Not measuredOnly Claude Sonnet 4.5 measured
SimpleBench54.3%60.6%Claude Sonnet 5
Chess puzzles10.0%45.0%Claude Sonnet 5
Mystery game puzzles17.0%35.0%Claude Sonnet 5
EBR-bench2.4%Not measuredOnly Claude Sonnet 4.5 measured
VPCT39.8%Not measuredOnly Claude Sonnet 4.5 measured
LMArena Text: hard prompts14621461Claude Sonnet 4.5
LMArena Text: expert14831489Claude Sonnet 5
LMArena Text: hard prompts in English14711464Claude Sonnet 4.5
LMArena Vision: diagramsNot measured1293Only Claude Sonnet 5 measured
LMArena Vision: homeworkNot measured1311Only Claude Sonnet 5 measured
Kagi LLM Benchmark57.9%Not measuredOnly Claude Sonnet 4.5 measured
Neither model has a result on the other 3 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index06
Math benchmarks for Claude Sonnet 4.5 and Claude Sonnet 5
BenchmarkClaude Sonnet 4.5Claude Sonnet 5Leader
AIME88.2%Not measuredOnly Claude Sonnet 4.5 measured
MGSM (Vals AI)94.3%Not measuredOnly Claude Sonnet 4.5 measured
OTIS Mock AIME77.8%94.7%Claude Sonnet 5
FrontierMath, tiers 1 to 323.9%65.6%Claude Sonnet 5
FrontierMath, tier 42.4%29.3%Claude Sonnet 5
ProofBench19.0%77.0%Claude Sonnet 5
MATH level 597.7%Not measuredOnly Claude Sonnet 4.5 measured
LMArena Text: math14491467Claude Sonnet 5
LMArena Text: mathematical work14621481Claude Sonnet 5
Neither model has a result on the other 4 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index38
Knowledge benchmarks for Claude Sonnet 4.5 and Claude Sonnet 5
BenchmarkClaude Sonnet 4.5Claude Sonnet 5Leader
MMLU Pro87.4%87.5%Claude Sonnet 5
LegalBench (Vals AI)84.1%83.9%Claude Sonnet 4.5
CaseLaw (Vals AI)62.2%Not measuredOnly Claude Sonnet 4.5 measured
CorpFin (Vals AI)62.0%67.0%Claude Sonnet 5
TaxEval (Vals AI)73.3%75.6%Claude Sonnet 5
MortgageTax (Vals AI)64.0%70.0%Claude Sonnet 5
MedQA (Vals AI)94.7%Not measuredOnly Claude Sonnet 4.5 measured
MedCode (Vals AI)44.1%47.5%Claude Sonnet 5
SimpleQA Verified30.7%33.7%Claude Sonnet 5
LMArena Text: science14481460Claude Sonnet 5
LMArena Text: medicine and healthcare14441442Claude Sonnet 4.5
LMArena Text: legal and government14511454Claude Sonnet 5
LMArena Text: business and finance14401436Claude Sonnet 4.5
FACTS Benchmark Suite40.6%Not measuredOnly Claude Sonnet 4.5 measured
Neither model has a result on the other 4 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index08
Web design benchmarks for Claude Sonnet 4.5 and Claude Sonnet 5
BenchmarkClaude Sonnet 4.5Claude Sonnet 5Leader
Design Arena: websites11891274Claude Sonnet 5
Design Arena: UI components11701277Claude Sonnet 5
Design Arena: game dev11661277Claude Sonnet 5
Design Arena: data viz11671251Claude Sonnet 5
Design Arena: 3D11571261Claude Sonnet 5
Design Arena: all code categories11821275Claude Sonnet 5
Design Arena: SVG11151192Claude Sonnet 5
Design Arena: ASCII art12051207Claude Sonnet 5
LiveBench designNot measured82.0%Only Claude Sonnet 5 measured
GeneralLike-for-likePublic boards, shown as published, kept out of the index913
General benchmarks for Claude Sonnet 4.5 and Claude Sonnet 5
BenchmarkClaude Sonnet 4.5Claude Sonnet 5Leader
LiveBench data analysisNot measured71.7%Only Claude Sonnet 5 measured
Public Benefits Bench (Vals AI)Not measured66.0%Only Claude Sonnet 5 measured
SAGE (Vals AI)36.1%48.9%Claude Sonnet 5
MedScribe (Vals AI)84.5%76.0%Claude Sonnet 4.5
Vals Multimodal IndexNot measured68.8%Only Claude Sonnet 5 measured
Long context reasoning (AA-LCR)Not measured82.0%Only Claude Sonnet 5 measured
DTBench83.2%92.5%Claude Sonnet 5
LMCA38.8%50.0%Claude Sonnet 5
Surface Evolver BenchNot measured60.0%Only Claude Sonnet 5 measured
LMArena Text14381442Claude Sonnet 5
LMArena Text: instruction following14591452Claude Sonnet 4.5
LMArena Text: creative writing14421416Claude Sonnet 4.5
LMArena VisionNot measured1273Only Claude Sonnet 5 measured
LMArena Text: English14501450Claude Sonnet 4.5
LMArena Text: non-English14251431Claude Sonnet 5
LMArena Text: Chinese14601478Claude Sonnet 5
LMArena Text: German14271442Claude Sonnet 5
LMArena Text: Japanese13901422Claude Sonnet 5
LMArena Text: Korean14021412Claude Sonnet 5
LMArena Text: French14591463Claude Sonnet 5
LMArena Text: Spanish14581438Claude Sonnet 4.5
LMArena Text: Russian14371450Claude Sonnet 5
LMArena Text: Polish14231425Claude Sonnet 5
LMArena Text: multi-turn14651453Claude Sonnet 4.5
LMArena Text: longer queries14751462Claude Sonnet 4.5
LMArena Text: writing and language14421436Claude Sonnet 4.5
LMArena Text: entertainment and media14281411Claude Sonnet 4.5
LMArena Vision: EnglishNot measured1275Only Claude Sonnet 5 measured
LMArena Vision: ChineseNot measured1331Only Claude Sonnet 5 measured
LMArena Vision: reading textNot measured1288Only Claude Sonnet 5 measured
LMArena Vision: humorNot measured1261Only Claude Sonnet 5 measured
LMArena Vision: creative writingNot measured1246Only Claude Sonnet 5 measured
LMArena Document14501466Claude Sonnet 5
Neither model has a result on the other 13 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Claude Sonnet 4.5 and Claude Sonnet 5
BenchmarkClaude Sonnet 4.5Claude Sonnet 5Leader
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 Sonnet 4.5$0.011Fits in one request
Claude Sonnet 5$0.0070Fits in one request

Claude Sonnet 5 costs less

Repository review

50K input + 3K output tokens

Claude Sonnet 4.5$0.195Fits in one request
Claude Sonnet 5$0.130Fits in one request

Claude Sonnet 5 costs less

Agent loop

200K input + 10K output tokens

Claude Sonnet 4.5$0.750Fits in one request
Claude Sonnet 5$0.500Fits in one request

Claude Sonnet 5 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 Sonnet 4.5 and Claude Sonnet 5
SpecificationAnthropic Claude Sonnet 4.5Anthropic Claude Sonnet 5
MakerAnthropicAnthropic
WeightsNo public weights listedNo public weights listed
ReleasedSeptember 29, 2025June 30, 2026
Context window1M1M
Input $/Mtok$3.00$2.00
Output $/Mtok$15.00$10.00
Coding index39.557.7
Independent measurements behind it36
Cost per solved task$1.16$1.87
SourceOpenRouter listingOpenRouter listing

Questions

Is Claude Sonnet 4.5 better than Claude Sonnet 5 for coding?

Pick Claude Sonnet 5 for coding: it scores 57.7 on the Vibe Coding Index against 39.5 for Claude Sonnet 4.5. Pick Claude Sonnet 4.5 for a lower cost per solved task ($1.16 against $1.87) and work measured by LMArena Text: coding or LMArena Text: hard prompts, where it leads.

Which is cheaper, Claude Sonnet 4.5 or Claude Sonnet 5?

Claude Sonnet 4.5 costs $3.00 per million input tokens and $15.00 per million output tokens; Claude Sonnet 5 costs $2.00 and $10.00. Per solved SWE-bench Verified task, Claude Sonnet 4.5 costs $1.16 and Claude Sonnet 5 $1.87.

Which has the bigger context window, Claude Sonnet 4.5 or Claude Sonnet 5?

Neither: both take 1M tokens.

How many benchmarks is the Claude Sonnet 4.5 vs Claude Sonnet 5 comparison based on?

74 benchmarks have a result for both models as of October 22, 2026. Claude Sonnet 4.5's coding index rests on 3 independent benchmarks and Claude Sonnet 5's on 6. 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-sonnet-4-5-vs-claude-sonnet-5" title="Claude Sonnet 4.5 vs Claude Sonnet 5, 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-sonnet-4-5-vs-claude-sonnet-5">LLMBenchmark.ai, Claude Sonnet 4.5 vs Claude Sonnet 5</a></p>