Noometry

Model comparison

Claude Sonnet 5 vs DeepSeek-V3.2-Exp

Claude Sonnet 5 is the stronger model overall, scoring 54.6 to 44.3 on the Noometry Index. DeepSeek-V3.2-Exp costs 14× less per token, which makes it the better buy when Claude Sonnet 5's lead doesn't matter for your workload.

Last verified . 32 shared benchmarks.

Claude Sonnet 5 Anthropic

54.6

Rank #29 Confirmed

DeepSeek-V3.2-Exp DeepSeek

44.3

Rank #78 Confirmed

Summary

  • They share 32 benchmarks with published results for both. Claude Sonnet 5 scores higher in 8 categories and DeepSeek-V3.2-Exp in 1 category; 9 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Claude Sonnet 5 leads 49.1 to 22.1.
  • The biggest single-benchmark swing is ProofBench: 77% for Claude Sonnet 5 and 8% for DeepSeek-V3.2-Exp.
  • DeepSeek-V3.2-Exp is cheaper at $0.26 / $0.38 per million input/output tokens, against $2 / $10 for Claude Sonnet 5.
  • Claude Sonnet 5 accepts more context: 1M tokens versus 164K.
  • DeepSeek-V3.2-Exp has downloadable open weights; the other is API-only.

Side by side

Claude Sonnet 5DeepSeek-V3.2-Exp
ProviderAnthropicDeepSeek
Noometry Index54.644.3
Released2026-06-292025-09-29
WeightsProprietaryOpen
Context window1M164K
Max output128K66K
Input $ / M tokens$2$0.26
Output $ / M tokens$10$0.38
Results tracked5149

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

Category by category

Coding Claude Sonnet 5 leads

Claude Sonnet 5: 55.5 (#26), DeepSeek-V3.2-Exp: 46.5 (#65)

BenchmarkClaude Sonnet 5DeepSeek-V3.2-Exp
LMArena WebDev15411362
SciCode54.3%38.9%
WeirdML68.8%39.5%
LMArena Coding14831454
DeepSWE53.8%—
FrontierCode42.7%—
SWE-bench Verified (bash only)—70%
Aider Polyglot—74.2%
CursorBench34.1%—
SWE-bench Multilingual—59%
GSO37.3%—
ALE-Bench1,463—

Agentic & Tool Use Claude Sonnet 5 leads

Claude Sonnet 5: 42.8 (#18), DeepSeek-V3.2-Exp: 32.7 (#59)

BenchmarkClaude Sonnet 5DeepSeek-V3.2-Exp
APEX-Agents54.5%21.3%
Vending-Bench 26,3781,034
Terminal-Bench—39.6%
Berkeley Function Calling Leaderboard—56.7%
TheAgentCompany—42.9%
GBAEval65.3%—
LMArena Search1194—

Reasoning Claude Sonnet 5 leads

Claude Sonnet 5: 49.1 (#39), DeepSeek-V3.2-Exp: 22.1 (#208)

BenchmarkClaude Sonnet 5DeepSeek-V3.2-Exp
NYT Connections (extended)75.1%36.7%
CritPt16.9%2.9%
Chess Puzzles35%14%
LMArena Hard Prompts14611434
DTBench92.5%87.7%
LMCA50%29.1%
Epoch Capabilities Index156.21146.27
ARC-AGI-2—4%
SimpleBench60.6%—
Kagi LLM Benchmark—52.2%
ARC-AGI-1—57%
Thematic Generalization—65%
Mystery Game Puzzles35%—
Surface Evolver Bench60%—
Bench to the Future 30.14—
ForecastBench61.1—

Math Claude Sonnet 5 leads

Claude Sonnet 5: 66.2 (#27), DeepSeek-V3.2-Exp: 41.7 (#87)

BenchmarkClaude Sonnet 5DeepSeek-V3.2-Exp
OTIS Mock AIME 2024-202594.7%87.8%
ProofBench77%8%
LMArena Math14671435
FrontierMath (Tiers 1-3)65.6%—
FrontierMath Tier 429.3%—
MathArena Final-Answer Competitions—57.7%
FrontierMath (Feb 2025 set)—22.1%
FrontierMath Tier 4 (v1)—2.1%

Knowledge Claude Sonnet 5 leads

Claude Sonnet 5: 55.6 (#47), DeepSeek-V3.2-Exp: 51.7 (#66)

BenchmarkClaude Sonnet 5DeepSeek-V3.2-Exp
GPQA Diamond90.5%83.4%
LMArena Expert14901436
SimpleQA Verified33.7%—
Vectara Hallucination Rate—5.3%

Multimodal Not comparable

Claude Sonnet 5: 42.4 (#31), DeepSeek-V3.2-Exp: —

BenchmarkClaude Sonnet 5DeepSeek-V3.2-Exp
LMArena Vision1274—
Blueprint-Bench 224.9%—
LMArena Document1466—

Multilingual Claude Sonnet 5 leads

Claude Sonnet 5: 53.8 (#55), DeepSeek-V3.2-Exp: 52.2 (#90)

BenchmarkClaude Sonnet 5DeepSeek-V3.2-Exp
LMArena Non-English14311409
LMArena Chinese14771461
LMArena French14601433
LMArena German14401440
LMArena Japanese14221374
LMArena Korean14111371
LMArena Russian14511424
LMArena Spanish14371440

Instruction Following Claude Sonnet 5 leads

Claude Sonnet 5: 76.3 (#41), DeepSeek-V3.2-Exp: 74.5 (#93)

BenchmarkClaude Sonnet 5DeepSeek-V3.2-Exp
LMArena Instruction Following14521413

Long Context DeepSeek-V3.2-Exp leads

Claude Sonnet 5: 44.8 (#55), DeepSeek-V3.2-Exp: 47.6 (#16)

BenchmarkClaude Sonnet 5DeepSeek-V3.2-Exp
LMArena Longer Query14631428
Fiction.LiveBench—83.3%
CL-bench—13.2%
CL-bench Life—9.5%

Writing & Preference Claude Sonnet 5 leads

Claude Sonnet 5: 69.2 (#25), DeepSeek-V3.2-Exp: 62.4 (#77)

BenchmarkClaude Sonnet 5DeepSeek-V3.2-Exp
LMArena Text14421425
LMArena Creative Writing14161403
EQ-Bench Creative Writing17941515
LMArena Multi-Turn14541427
EQ-Bench 41236—

Frequently asked questions

Is Claude Sonnet 5 better than DeepSeek-V3.2-Exp?

Claude Sonnet 5 is the stronger model overall, scoring 54.6 to 44.3 on the Noometry Index. DeepSeek-V3.2-Exp costs 14× less per token, which makes it the better buy when Claude Sonnet 5's lead doesn't matter for your workload.

Which is cheaper, Claude Sonnet 5 or DeepSeek-V3.2-Exp?

DeepSeek-V3.2-Exp is cheaper. It lists at $0.26 per million input tokens and $0.38 per million output tokens; Claude Sonnet 5 lists at $2 and $10.

Is Claude Sonnet 5 or DeepSeek-V3.2-Exp better for coding?

Claude Sonnet 5 scores higher on coding benchmarks: 55.5 versus 46.5 in the Noometry coding category.

Which has the bigger context window?

Claude Sonnet 5 does, with 1M tokens against 164K.

How many benchmarks do Claude Sonnet 5 and DeepSeek-V3.2-Exp share?

32 benchmarks have published results for both models. Claude Sonnet 5 has 51 scored results on Noometry and DeepSeek-V3.2-Exp has 49.

Related comparisons

Go deeper