Noometry

Model comparison

Claude Sonnet 4.5 vs DeepSeek-V3

Claude Sonnet 4.5 is the stronger model overall, scoring 44.1 to 39.5 on the Noometry Index. DeepSeek-V3 costs 15× less per token, which makes it the better buy when Claude Sonnet 4.5's lead doesn't matter for your workload.

Last verified . 38 shared benchmarks.

Claude Sonnet 4.5 Anthropic

44.1

Rank #81 Confirmed

DeepSeek-V3 DeepSeek

39.5

Rank #166 Confirmed

Summary

  • They share 38 benchmarks with published results for both. Claude Sonnet 4.5 scores higher in 8 categories and DeepSeek-V3 in 0 categories; 7 gaps are clear of the uncertainty.
  • The widest gap is in long context, where Claude Sonnet 4.5 leads 45.2 to 34.0.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 77.8% for Claude Sonnet 4.5 and 37.8% for DeepSeek-V3.
  • DeepSeek-V3 is cheaper at $0.24 / $0.90 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.5.
  • Claude Sonnet 4.5 accepts more context: 200K tokens versus 164K.
  • DeepSeek-V3 has downloadable open weights; the other is API-only.

Side by side

Claude Sonnet 4.5DeepSeek-V3
ProviderAnthropicDeepSeek
Noometry Index44.139.5
Released2025-09-292024-12-26
WeightsProprietaryOpen
Context window200K164K
Max output64K164K
Input $ / M tokens$3$0.24
Output $ / M tokens$15$0.90
Results tracked7360

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), DeepSeek-V3: 42.3 (#106)

BenchmarkClaude Sonnet 4.5DeepSeek-V3
SciCode44.7%35.8%
WeirdML47.7%36.1%
LMArena Coding14891368
SWE-bench Verified71.3%—
SWE-bench Verified (bash only)71.4%—
Aider Polyglot—55.1%
LMArena WebDev1393—
SWE-bench Multilingual67%—
GSO14.7%—
BigCodeBench Instruct—50%
LiveBench Coding—70.9%
BigCodeBench Complete—62.2%
ALE-Bench796.15—
AlgoTune1.52—
HumanEval+—86.6%
MBPP+—73%

Agentic & Tool Use Not comparable

Claude Sonnet 4.5: 38.3 (#32), DeepSeek-V3: —

BenchmarkClaude Sonnet 4.5DeepSeek-V3
METR Time Horizons67.4%49.6%
Terminal-Bench46.5%—
Berkeley Function Calling Leaderboard73.2%—
GDPval42.5%—
Remote Labor Index2.1%—
τ²-bench Airline72%—
τ²-bench Banking25.3%—
τ²-bench Retail72.4%—
τ²-bench Telecom84.9%—
Cybench60%—
DeepResearch Bench52.6%—
OSWorld62.9%—
LMArena Search1159—
Vending-Bench 23,839—

Reasoning Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 26.9 (#125), DeepSeek-V3: 20.5 (#236)

BenchmarkClaude Sonnet 4.5DeepSeek-V3
SimpleBench54.3%27.2%
Kagi LLM Benchmark57.9%52.3%
CritPt1.1%0%
LMArena Hard Prompts14621365
DTBench83.2%64.8%
LMCA38.8%15.5%
Epoch Capabilities Index146.84135.94
ForecastBench61.959.1
ARC-AGI-213.6%—
NYT Connections (extended)37.3%—
ARC-AGI-163.7%—
Chess Puzzles12%—
EnigmaEval6%—
EBR-Bench2.4%—
LiveBench Reasoning—65.8%
Mystery Game Puzzles17%—
LiveBench Data Analysis—60.9%
BIG-Bench Hard—87.5%
HellaSwag—88.9%
LiveBench—66.9%
PIQA—84.7%
WinoGrande—85.2%

Math Too close to call

Claude Sonnet 4.5: 32.3 (#216), DeepSeek-V3: 32.1 (#219)

BenchmarkClaude Sonnet 4.5DeepSeek-V3
OTIS Mock AIME 2024-202577.8%37.8%
Omni-MATH55.3%40.3%
LMArena Math14491373
MATH Level 597.7%75.5%
FrontierMath (Feb 2025 set)15.2%1.7%
FrontierMath (Tiers 1-3)23.9%—
FrontierMath Tier 42.4%—
ProofBench19%—
LiveBench Math—73.5%
FrontierMath Tier 4 (v1)4.2%—

Knowledge Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 48.4 (#76), DeepSeek-V3: 37.5 (#155)

BenchmarkClaude Sonnet 4.5DeepSeek-V3
GPQA Diamond82.3%67.6%
MMLU-Pro86.9%72.3%
Vectara Hallucination Rate12%6.1%
GPQA (HELM)68.6%53.8%
LMArena Expert14821351
Humanity's Last Exam13.7%—
SimpleQA Verified30.7%—
Confabulations—26.1%
ARC (AI2) Challenge—95.3%
MMLU—87.2%
TriviaQA—82.9%

Multimodal Not comparable

Claude Sonnet 4.5: 34.8 (#89), DeepSeek-V3: —

BenchmarkClaude Sonnet 4.5DeepSeek-V3
VPCT39.8%—
LMArena Document1450—

Multilingual Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 53.4 (#69), DeepSeek-V3: 48.5 (#143)

BenchmarkClaude Sonnet 4.5DeepSeek-V3
LMArena Non-English14251358
LMArena Chinese14591391
LMArena French14581385
LMArena German14271374
LMArena Japanese13901333
LMArena Korean14031319
LMArena Russian14371373
LMArena Spanish14571358

Instruction Following Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 75.0 (#78), DeepSeek-V3: 72.8 (#130)

BenchmarkClaude Sonnet 4.5DeepSeek-V3
IFEval85%83.2%
LMArena Instruction Following14591345
LiveBench Instruction Following—81.5%

Long Context Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 45.2 (#46), DeepSeek-V3: 34.0 (#253)

BenchmarkClaude Sonnet 4.5DeepSeek-V3
LMArena Longer Query14761352
Fiction.LiveBench—50%

Writing & Preference Claude Sonnet 4.5 leads

Claude Sonnet 4.5: 66.5 (#34), DeepSeek-V3: 57.4 (#130)

BenchmarkClaude Sonnet 4.5DeepSeek-V3
LMArena Text14391375
LMArena Creative Writing14421364
EQ-Bench Creative Writing16781472
WildBench85.4%83%
LMArena Multi-Turn14651389
Short-Story Creative Writing—77%
LiveBench Language—49.1%

Frequently asked questions

Is Claude Sonnet 4.5 better than DeepSeek-V3?

Claude Sonnet 4.5 is the stronger model overall, scoring 44.1 to 39.5 on the Noometry Index. DeepSeek-V3 costs 15× less per token, which makes it the better buy when Claude Sonnet 4.5's lead doesn't matter for your workload.

Which is cheaper, Claude Sonnet 4.5 or DeepSeek-V3?

DeepSeek-V3 is cheaper. It lists at $0.24 per million input tokens and $0.90 per million output tokens; Claude Sonnet 4.5 lists at $3 and $15.

Is Claude Sonnet 4.5 or DeepSeek-V3 better for coding?

Claude Sonnet 4.5 scores higher on coding benchmarks: 47.3 versus 42.3 in the Noometry coding category.

Which has the bigger context window?

Claude Sonnet 4.5 does, with 200K tokens against 164K.

How many benchmarks do Claude Sonnet 4.5 and DeepSeek-V3 share?

38 benchmarks have published results for both models. Claude Sonnet 4.5 has 73 scored results on Noometry and DeepSeek-V3 has 60.

Related comparisons

Go deeper