Noometry

Model comparison

Claude Opus 5 vs DeepSeek-V3

Claude Opus 5 is the stronger model overall, scoring 67.8 to 39.5 on the Noometry Index. DeepSeek-V3 costs 25× less per token, which makes it the better buy when Claude Opus 5's lead doesn't matter for your workload.

Last verified . 27 shared benchmarks.

Claude Opus 5 Anthropic

67.8

Rank #4 Confirmed

DeepSeek-V3 DeepSeek

39.5

Rank #166 Confirmed

Summary

  • They share 27 benchmarks with published results for both. Claude Opus 5 scores higher in 8 categories and DeepSeek-V3 in 0 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Claude Opus 5 leads 77.2 to 20.5.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 98.9% for Claude Opus 5 and 37.8% for DeepSeek-V3.
  • DeepSeek-V3 is cheaper at $0.24 / $0.90 per million input/output tokens, against $5 / $25 for Claude Opus 5.
  • Claude Opus 5 accepts more context: 1M tokens versus 164K.
  • DeepSeek-V3 has downloadable open weights; the other is API-only.

Side by side

Claude Opus 5DeepSeek-V3
ProviderAnthropicDeepSeek
Noometry Index67.839.5
Released2026-07-242024-12-26
WeightsProprietaryOpen
Context window1M164K
Max output128K164K
Input $ / M tokens$5$0.24
Output $ / M tokens$25$0.90
Results tracked5760

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

Category by category

Coding Claude Opus 5 leads

Claude Opus 5: 67.5 (#5), DeepSeek-V3: 42.3 (#106)

BenchmarkClaude Opus 5DeepSeek-V3
SciCode56.4%35.8%
WeirdML91.8%36.1%
LMArena Coding15341368
DeepSWE73.6%—
FrontierCode53.4%—
Aider Polyglot—55.1%
CursorBench46.6%—
LMArena WebDev1691—
FrontierSWE52%—
BigCodeBench Instruct—50%
LiveBench Coding—70.9%
BigCodeBench Complete—62.2%
ALE-Bench2,165—
HumanEval+—86.6%
MBPP+—73%

Agentic & Tool Use Not comparable

Claude Opus 5: 55.6 (#1), DeepSeek-V3: —

BenchmarkClaude Opus 5DeepSeek-V3
APEX-Agents65.8%—
OSWorld 2.031.4%—
τ²-bench Banking48.7%—
PostTrainBench35%—
BALROG63.4%—
GBAEval79.6%—
GDP.pdf24%—
METR Time Horizons—49.6%
Vending-Bench 211,182—

Reasoning Claude Opus 5 leads

Claude Opus 5: 77.2 (#4), DeepSeek-V3: 20.5 (#236)

BenchmarkClaude Opus 5DeepSeek-V3
SimpleBench80.6%27.2%
CritPt29.1%0%
LMArena Hard Prompts15261365
DTBench97.9%64.8%
LMCA64.5%15.5%
Epoch Capabilities Index162.78135.94
ARC-AGI-290.4%—
Kagi LLM Benchmark—52.3%
NYT Connections (extended)94.3%—
ARC-AGI-197.5%—
Chess Puzzles42%—
EBR-Bench45.7%—
LiveBench Reasoning—65.8%
Mystery Game Puzzles59%—
LiveBench Data Analysis—60.9%
Bench to the Future 30.12—
BIG-Bench Hard—87.5%
ForecastBench—59.1
HellaSwag—88.9%
LiveBench—66.9%
PIQA—84.7%
WinoGrande—85.2%

Math Claude Opus 5 leads

Claude Opus 5: 86.2 (#8), DeepSeek-V3: 32.1 (#219)

BenchmarkClaude Opus 5DeepSeek-V3
OTIS Mock AIME 2024-202598.9%37.8%
LMArena Math15311373
FrontierMath (Tiers 1-3)85.6%—
FrontierMath Tier 473.2%—
ProofBench99%—
Omni-MATH—40.3%
LiveBench Math—73.5%
MATH Level 5—75.5%
FrontierMath (Feb 2025 set)—1.7%

Knowledge Claude Opus 5 leads

Claude Opus 5: 66.8 (#9), DeepSeek-V3: 37.5 (#155)

BenchmarkClaude Opus 5DeepSeek-V3
GPQA Diamond93.9%67.6%
LMArena Expert15571351
SimpleQA Verified59.9%—
MMLU-Pro—72.3%
Confabulations—26.1%
Vectara Hallucination Rate—6.1%
GPQA (HELM)—53.8%
ARC (AI2) Challenge—95.3%
MMLU—87.2%
TriviaQA—82.9%

Multimodal Not comparable

Claude Opus 5: 50.8 (#8), DeepSeek-V3: —

BenchmarkClaude Opus 5DeepSeek-V3
LMArena Vision1319—
Blueprint-Bench 230.4%—
Furniture Assembly60.8%—
LMArena Document1516—

Multilingual Claude Opus 5 leads

Claude Opus 5: 58.8 (#4), DeepSeek-V3: 48.5 (#143)

BenchmarkClaude Opus 5DeepSeek-V3
LMArena Non-English15011358
LMArena Chinese15741391
LMArena French15191385
LMArena German15241374
LMArena Japanese15161333
LMArena Korean15211319
LMArena Russian15071373
LMArena Spanish15191358

Instruction Following Claude Opus 5 leads

Claude Opus 5: 79.2 (#7), DeepSeek-V3: 72.8 (#130)

BenchmarkClaude Opus 5DeepSeek-V3
LMArena Instruction Following15171345
LiveBench Instruction Following—81.5%
IFEval—83.2%

Long Context Claude Opus 5 leads

Claude Opus 5: 46.5 (#21), DeepSeek-V3: 34.0 (#253)

BenchmarkClaude Opus 5DeepSeek-V3
LMArena Longer Query15151352
Fiction.LiveBench—50%

Writing & Preference Claude Opus 5 leads

Claude Opus 5: 79.2 (#1), DeepSeek-V3: 57.4 (#130)

BenchmarkClaude Opus 5DeepSeek-V3
LMArena Text15071375
LMArena Creative Writing14911364
EQ-Bench Creative Writing21331472
LMArena Multi-Turn14991389
Short-Story Creative Writing—77%
WildBench—83%
EQ-Bench 41385—
LiveBench Language—49.1%

Frequently asked questions

Is Claude Opus 5 better than DeepSeek-V3?

Claude Opus 5 is the stronger model overall, scoring 67.8 to 39.5 on the Noometry Index. DeepSeek-V3 costs 25× less per token, which makes it the better buy when Claude Opus 5's lead doesn't matter for your workload.

Which is cheaper, Claude Opus 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 Opus 5 lists at $5 and $25.

Is Claude Opus 5 or DeepSeek-V3 better for coding?

Claude Opus 5 scores higher on coding benchmarks: 67.5 versus 42.3 in the Noometry coding category.

Which has the bigger context window?

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

How many benchmarks do Claude Opus 5 and DeepSeek-V3 share?

27 benchmarks have published results for both models. Claude Opus 5 has 57 scored results on Noometry and DeepSeek-V3 has 60.

Related comparisons

Go deeper