Noometry

Model comparison

Claude Sonnet 4.5 vs Claude Sonnet 4.6

Claude Sonnet 4.6 is the stronger model overall, scoring 50.3 to 44.1 on the Noometry Index.

Last verified . 46 shared benchmarks.

Claude Sonnet 4.5 Anthropic

44.1

Rank #81 Confirmed

Claude Sonnet 4.6 Anthropic

50.3

Rank #50 Confirmed

Summary

  • They share 46 benchmarks with published results for both. Claude Sonnet 4.5 scores higher in 1 category and Claude Sonnet 4.6 in 9 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in math, where Claude Sonnet 4.6 leads 52.9 to 32.3.
  • The biggest single-benchmark swing is ARC-AGI-2: 13.6% for Claude Sonnet 4.5 and 60.4% for Claude Sonnet 4.6.
  • Both cost about the same: $3 input and $15 output per million tokens.
  • Claude Sonnet 4.6 accepts more context: 1M tokens versus 200K.

Side by side

Claude Sonnet 4.5Claude Sonnet 4.6
ProviderAnthropicAnthropic
Noometry Index44.150.3
Released2025-09-292026-02-17
WeightsProprietaryProprietary
Context window200K1M
Max output64K128K
Input $ / M tokens$3$3
Output $ / M tokens$15$15
Results tracked7357

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

Category by category

Coding Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 47.3 (#61), Claude Sonnet 4.6: 46.3 (#67)

BenchmarkClaude Sonnet 4.5Claude Sonnet 4.6
SWE-bench Verified71.3%75.2%
LMArena WebDev13931522
SciCode44.7%46.8%
WeirdML47.7%66.1%
LMArena Coding14891504
ALE-Bench796.151,327
DeepSWE—29.9%
FrontierCode—24.3%
SWE-bench Verified (bash only)71.4%—
SWE-bench Multilingual67%—
GSO14.7%—
AlgoTune1.52—

Agentic & Tool Use Too close to call

Claude Sonnet 4.5: 38.3 (#32), Claude Sonnet 4.6: 39.1 (#28)

BenchmarkClaude Sonnet 4.5Claude Sonnet 4.6
Terminal-Bench46.5%53.4%
DeepResearch Bench52.6%54.9%
OSWorld62.9%72.1%
LMArena Search11591221
Vending-Bench 23,8397,204
APEX-Agents—43%
Berkeley Function Calling Leaderboard73.2%—
OSWorld 2.0—9.3%
GDPval42.5%—
Remote Labor Index2.1%—
τ²-bench Airline72%—
τ²-bench Banking25.3%—
τ²-bench Retail72.4%—
τ²-bench Telecom84.9%—
Cybench60%—
ExploitBench—23.6%
GBAEval—48.8%
GDP.pdf—18%
METR Time Horizons67.4%—

Reasoning Claude Sonnet 4.6 leads

Claude Sonnet 4.5: 26.9 (#125), Claude Sonnet 4.6: 46.1 (#45)

BenchmarkClaude Sonnet 4.5Claude Sonnet 4.6
ARC-AGI-213.6%60.4%
NYT Connections (extended)37.3%80.9%
ARC-AGI-163.7%86.5%
CritPt1.1%3.1%
Chess Puzzles12%13%
LMArena Hard Prompts14621484
Mystery Game Puzzles17%16%
DTBench83.2%89.9%
LMCA38.8%46.5%
Epoch Capabilities Index146.84152.24
ForecastBench61.962
SimpleBench54.3%—
Kagi LLM Benchmark57.9%—
EnigmaEval6%—
Thematic Generalization—76.3%
EBR-Bench2.4%—

Math Claude Sonnet 4.6 leads

Claude Sonnet 4.5: 32.3 (#216), Claude Sonnet 4.6: 52.9 (#49)

BenchmarkClaude Sonnet 4.5Claude Sonnet 4.6
OTIS Mock AIME 2024-202577.8%85.8%
ProofBench19%45%
LMArena Math14491462
FrontierMath (Feb 2025 set)15.2%32.4%
FrontierMath Tier 4 (v1)4.2%8.3%
FrontierMath (Tiers 1-3)23.9%—
FrontierMath Tier 42.4%—
Omni-MATH55.3%—
MATH Level 597.7%—

Knowledge Claude Sonnet 4.6 leads

Claude Sonnet 4.5: 48.4 (#76), Claude Sonnet 4.6: 51.7 (#65)

BenchmarkClaude Sonnet 4.5Claude Sonnet 4.6
GPQA Diamond82.3%87.4%
SimpleQA Verified30.7%35.5%
Vectara Hallucination Rate12%10.6%
LMArena Expert14821500
Humanity's Last Exam13.7%—
MMLU-Pro86.9%—
GPQA (HELM)68.6%—

Multimodal Claude Sonnet 4.6 leads

Claude Sonnet 4.5: 34.8 (#89), Claude Sonnet 4.6: 38.0 (#68)

BenchmarkClaude Sonnet 4.5Claude Sonnet 4.6
LMArena Document14501482
LMArena Vision—1283
VPCT39.8%—
Blueprint-Bench 2—6.7%

Multilingual Claude Sonnet 4.6 leads

Claude Sonnet 4.5: 53.4 (#69), Claude Sonnet 4.6: 54.4 (#41)

BenchmarkClaude Sonnet 4.5Claude Sonnet 4.6
LMArena Non-English14251440
LMArena Chinese14591491
LMArena French14581465
LMArena German14271428
LMArena Japanese13901420
LMArena Korean14031411
LMArena Russian14371440
LMArena Spanish14571464

Instruction Following Claude Sonnet 4.6 leads

Claude Sonnet 4.5: 75.0 (#78), Claude Sonnet 4.6: 77.4 (#25)

BenchmarkClaude Sonnet 4.5Claude Sonnet 4.6
LMArena Instruction Following14591475
IFEval85%—

Long Context Too close to call

Claude Sonnet 4.5: 45.2 (#46), Claude Sonnet 4.6: 45.3 (#44)

BenchmarkClaude Sonnet 4.5Claude Sonnet 4.6
LMArena Longer Query14761479

Writing & Preference Claude Sonnet 4.6 leads

Claude Sonnet 4.5: 66.5 (#34), Claude Sonnet 4.6: 70.2 (#22)

BenchmarkClaude Sonnet 4.5Claude Sonnet 4.6
LMArena Text14391458
LMArena Creative Writing14421435
EQ-Bench Creative Writing16781810
LMArena Multi-Turn14651464
WildBench85.4%—
EQ-Bench 4—1207

Frequently asked questions

Is Claude Sonnet 4.5 better than Claude Sonnet 4.6?

Claude Sonnet 4.6 is the stronger model overall, scoring 50.3 to 44.1 on the Noometry Index.

Which is cheaper, Claude Sonnet 4.5 or Claude Sonnet 4.6?

Claude Sonnet 4.6 is cheaper. It lists at $3 per million input tokens and $15 per million output tokens; Claude Sonnet 4.5 lists at $3 and $15.

Is Claude Sonnet 4.5 or Claude Sonnet 4.6 better for coding?

Claude Sonnet 4.5 scores higher on coding benchmarks: 47.3 versus 46.3 in the Noometry coding category.

Which has the bigger context window?

Claude Sonnet 4.6 does, with 1M tokens against 200K.

How many benchmarks do Claude Sonnet 4.5 and Claude Sonnet 4.6 share?

46 benchmarks have published results for both models. Claude Sonnet 4.5 has 73 scored results on Noometry and Claude Sonnet 4.6 has 57.

Related comparisons

Go deeper