Noometry

Model comparison

Claude Sonnet 4 vs Claude Sonnet 4.5

Claude Sonnet 4.5 is the stronger model overall, scoring 44.1 to 40.8 on the Noometry Index.

Last verified . 50 shared benchmarks.

Claude Sonnet 4 Anthropic

40.8

Rank #145 Confirmed

Claude Sonnet 4.5 Anthropic

44.1

Rank #81 Confirmed

Summary

  • They share 50 benchmarks with published results for both. Claude Sonnet 4 scores higher in 2 categories and Claude Sonnet 4.5 in 8 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in long context, where Claude Sonnet 4.5 leads 45.2 to 33.7.
  • The biggest single-benchmark swing is Cybench: 35% for Claude Sonnet 4 and 60% for Claude Sonnet 4.5.
  • Both cost about the same: $3 input and $15 output per million tokens.

Side by side

Claude Sonnet 4Claude Sonnet 4.5
ProviderAnthropicAnthropic
Noometry Index40.844.1
Released2025-05-222025-09-29
WeightsProprietaryProprietary
Context window200K200K
Max output64K64K
Input $ / M tokens$3$3
Output $ / M tokens$15$15
Results tracked5873

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Claude Sonnet 4.5 leads

Claude Sonnet 4: 43.5 (#88), Claude Sonnet 4.5: 47.3 (#61)

BenchmarkClaude Sonnet 4Claude Sonnet 4.5
SWE-bench Verified (bash only)64.9%71.4%
SciCode40%44.7%
GSO4.9%14.7%
WeirdML46.1%47.7%
LMArena Coding14141489
ALE-Bench655.35796.15
SWE-bench Verified—71.3%
Aider Polyglot61.3%—
LMArena WebDev—1393
SWE-bench Multilingual—67%
AlgoTune—1.52

Agentic & Tool Use Too close to call

Claude Sonnet 4: 38.5 (#31), Claude Sonnet 4.5: 38.3 (#32)

BenchmarkClaude Sonnet 4Claude Sonnet 4.5
Cybench35%60%
DeepResearch Bench46.6%52.6%
OSWorld43.9%62.9%
METR Time Horizons62%67.4%
Terminal-Bench—46.5%
Berkeley Function Calling Leaderboard—73.2%
GDPval—42.5%
Remote Labor Index—2.1%
TheAgentCompany33.1%—
τ²-bench Airline—72%
τ²-bench Banking—25.3%
τ²-bench Retail—72.4%
τ²-bench Telecom—84.9%
LMArena Search—1159
Vending-Bench 2—3,839

Reasoning Claude Sonnet 4.5 leads

Claude Sonnet 4: 22.9 (#187), Claude Sonnet 4.5: 26.9 (#125)

BenchmarkClaude Sonnet 4Claude Sonnet 4.5
ARC-AGI-25.9%13.6%
SimpleBench45.5%54.3%
Kagi LLM Benchmark73%57.9%
ARC-AGI-140%63.7%
CritPt0.3%1.1%
EnigmaEval3.1%6%
LMArena Hard Prompts13721462
DTBench77.1%83.2%
LMCA29%38.8%
Epoch Capabilities Index141.69146.84
ForecastBench60.261.9
NYT Connections (extended)—37.3%
Chess Puzzles—12%
EBR-Bench—2.4%
Mystery Game Puzzles—17%

Math Claude Sonnet 4 leads

Claude Sonnet 4: 43.3 (#80), Claude Sonnet 4.5: 32.3 (#216)

BenchmarkClaude Sonnet 4Claude Sonnet 4.5
OTIS Mock AIME 2024-202571.1%77.8%
Omni-MATH60.2%55.3%
LMArena Math13751449
MATH Level 584.4%97.7%
FrontierMath (Feb 2025 set)4.1%15.2%
FrontierMath Tier 4 (v1)0%4.2%
FrontierMath (Tiers 1-3)—23.9%
FrontierMath Tier 4—2.4%
ProofBench—19%

Knowledge Claude Sonnet 4.5 leads

Claude Sonnet 4: 41.8 (#108), Claude Sonnet 4.5: 48.4 (#76)

BenchmarkClaude Sonnet 4Claude Sonnet 4.5
GPQA Diamond79.2%82.3%
Humanity's Last Exam7.8%13.7%
MMLU-Pro84.3%86.9%
Vectara Hallucination Rate10.3%12%
GPQA (HELM)70.6%68.6%
LMArena Expert13721482
SimpleQA Verified—30.7%
Confabulations13.2%—

Multimodal Claude Sonnet 4.5 leads

Claude Sonnet 4: 26.2 (#121), Claude Sonnet 4.5: 34.8 (#89)

BenchmarkClaude Sonnet 4Claude Sonnet 4.5
VPCT34%39.8%
LMArena Vision1191—
GeoBench37%—
LMArena Document—1450
MindCube44.8%—

Multilingual Claude Sonnet 4.5 leads

Claude Sonnet 4: 46.7 (#156), Claude Sonnet 4.5: 53.4 (#69)

BenchmarkClaude Sonnet 4Claude Sonnet 4.5
LMArena Non-English13331425
LMArena Chinese13501459
LMArena French13631458
LMArena German13311427
LMArena Japanese13021390
LMArena Korean12911403
LMArena Russian13551437
LMArena Spanish13571457

Instruction Following Claude Sonnet 4.5 leads

Claude Sonnet 4: 71.7 (#145), Claude Sonnet 4.5: 75.0 (#78)

BenchmarkClaude Sonnet 4Claude Sonnet 4.5
IFEval84%85%
LMArena Instruction Following13761459

Long Context Claude Sonnet 4.5 leads

Claude Sonnet 4: 33.7 (#259), Claude Sonnet 4.5: 45.2 (#46)

BenchmarkClaude Sonnet 4Claude Sonnet 4.5
LMArena Longer Query13981476
Fiction.LiveBench46.9%—

Writing & Preference Claude Sonnet 4.5 leads

Claude Sonnet 4: 57.1 (#132), Claude Sonnet 4.5: 66.5 (#34)

BenchmarkClaude Sonnet 4Claude Sonnet 4.5
LMArena Text13511439
LMArena Creative Writing13451442
EQ-Bench Creative Writing14831678
WildBench83.8%85.4%
LMArena Multi-Turn13761465
Short-Story Creative Writing81.4%—

Frequently asked questions

Is Claude Sonnet 4 better than Claude Sonnet 4.5?

Claude Sonnet 4.5 is the stronger model overall, scoring 44.1 to 40.8 on the Noometry Index.

Which is cheaper, Claude Sonnet 4 or Claude Sonnet 4.5?

Claude Sonnet 4.5 is cheaper. It lists at $3 per million input tokens and $15 per million output tokens; Claude Sonnet 4 lists at $3 and $15.

Is Claude Sonnet 4 or Claude Sonnet 4.5 better for coding?

Claude Sonnet 4.5 scores higher on coding benchmarks: 47.3 versus 43.5 in the Noometry coding category.

Which has the bigger context window?

Both accept 200K tokens.

How many benchmarks do Claude Sonnet 4 and Claude Sonnet 4.5 share?

50 benchmarks have published results for both models. Claude Sonnet 4 has 58 scored results on Noometry and Claude Sonnet 4.5 has 73.

Related comparisons

Go deeper