DeepSeek V3 0324

March 25, 2025Last synced Jul 10, 2026, 8:59 PM

Human Frontier

82.5

Human-calibrated frontier signal, backed by Arena-style preference evidence and separate from raw AgMoBench benchmark composite scores.

Blended Price

$1.21/M

Input Price

$1.20/M

Output Price

$1.25/M

Speed

—

TTFT

—

Benchmark Scores

Intelligence Index

Performance Radar

DeepSeek V3 0324

Field Average

External Benchmarks

AIME 2024benchmark_matrix

How DeepSeek V3 0324 Compares

Axes

X Axis

Y Axis

Bubble Size

Filters

Blended Price (USD)$0.00 – $37.5

AgMoBench Overall≥ 1.9

Providers

ai2ai21-labsalibabaanthropicarceeawsazurebaidu

Show quadrants

google

mistral

deepseek

stepfun

openai

anthropic

liquidai

meta

cohere

nvidia

azure

kimi

minimax

alibaba

ibm

nous-research

aws

multiversecomputing

zai

deepcogito

kwaikat

xai

xiaomi

ai21-labs

inclusionai

baidu

sarvam

bytedance_seed

upstage

reka-ai

inception

arcee

tencent

ai2

perplexity

swiss-ai-initiative

Bubble size = Context Window

8191.00

2000000.00

Compare with other models

GPQA Diamond

0.7 / 100

LiveCodeBench

0.4 / 100

Terminal-Bench Hard

0.2 / 100

τ²-Bench

0.5 / 100

AIME 2025 (Matrix)benchmark_matrix

46.7

Arena-Hard Autobenchmark_matrix

39.9

Codeforces Ratingbenchmark_matrix

1650.0

GPQA Diamond (Matrix)benchmark_matrix

68.4

HLE (Matrix)benchmark_matrix

5.2

HMMT Feb 2025benchmark_matrix

27.5

HumanEvalbenchmark_matrix

85.0

IFEvalbenchmark_matrix

81.1

LiveCodeBench (Matrix)benchmark_matrix

49.2

MATH-500 (Matrix)benchmark_matrix

94.0

MMLUbenchmark_matrix

89.4

MMLU-Pro (Matrix)benchmark_matrix

81.2

SimpleQAbenchmark_matrix

27.7

simpleqabenchmark_matrix

27.7

SWE-bench Verifiedbenchmark_matrix

38.8 / 100

AA-Omniscience AccuracyPredicted

38.9 / 100

AA-Omniscience Hallucination RatePredicted

95.0 / 100

Aider PolyglotPredicted

38.8 / 100

AlpacaEval 2.0 LCPredicted

23.3 / 100

AlpacaEval 2.0 RawPredicted

16.3 / 100

ARC-AGI-1Predicted

5.1 / 100

ARC-AGI-1 Cost per TaskPredicted

0.0

ARC-AGI-2Predicted

0.0 / 100

ARC-AGI-2 Cost per TaskPredicted

0.1

BFCL (Berkeley Function Calling)Predicted

43.3

BigCodeBench CompletePredicted

59.4 / 100

BigCodeBench InstructPredicted

49.2 / 100

AA Intelligence Index (Matrix)Predicted

68.6

AA Long Context Reasoning (Matrix)Predicted

70.0

BrowseCompPredicted

75.7

BRUMO 2025Predicted

75.0

CMIMC 2025Predicted

67.5

HMMT Nov 2025Predicted

70.4

IFBench (Matrix)Predicted

36.4

IMO 2025Predicted

11.2

MathArena Apex 2025Predicted

0.5

MMMU-ProPredicted

81.4

MRCR v2Predicted

81.3

OSWorldPredicted

34.3

SMT 2025Predicted

76.8

SWE-bench ProPredicted

31.7

Tau-Bench Telecom (Matrix)Predicted

37.9

Terminal-Bench 2.0Predicted

22.0

Terminal-Bench 1.0Predicted

15.4

USAMO 2025Predicted

3.4

Video-MMUPredicted

70.1

browsecompPredicted

74.1

BullshitBenchPredicted

12.5 / 100

Apex AgentsPredicted

1.8

Arc Agi 2Predicted

0.0

BIG-Bench HardPredicted

3.0

CAD-EvalPredicted

5.0

Chess PuzzlesPredicted

0.1

CyBenchPredicted

0.1

DeepResearchBenchPredicted

0.3

GeoBenchPredicted

0.0

GSM8K (Epoch)Predicted

19.1

HellaSwagPredicted

10.2

PosttrainbenchPredicted

0.0

SimpleQA Verified (Epoch)Predicted

0.2

The Agent CompanyPredicted

1.7

TriviaQAPredicted

3.2

WinoGrandePredicted

0.9

FrontierMathPredicted

10.5 / 100

GAIA Level 1Predicted

55.8

GAIA Level 2Predicted

37.5

GAIA Level 3Predicted

6.3

GAIAPredicted

39.4 / 100

LegalBenchPredicted

89.4 / 100

LiveBench CodingPredicted

63.6 / 100

LiveBench Data AnalysisPredicted

38.6 / 100

LiveBench LanguagePredicted

56.5 / 100

LiveBench MathPredicted

55.6 / 100

LiveBench OverallPredicted

46.9 / 100

LiveBench ReasoningPredicted

43.2 / 100

LongBench v2 EasyPredicted

49.2

LongBench v2 HardPredicted

45.7

LongBench v2Predicted

41.2 / 100

MathVistaPredicted

66.3 / 100

MedQA (USMLE)Predicted

82.1

MLE-benchPredicted

16.7 / 100

MMMUPredicted

71.5 / 100

MMTU Table UnderstandingPredicted

56.7 / 100

MT-BenchPredicted

8.6 / 10

NoLiMa (NIAH)Predicted

95.5 / 100

OCRBench v2Predicted

83.9 / 100

Open LLM AveragePredicted

38.0 / 100

Open LLM: BBHPredicted

65.5 / 100

Open LLM: GPQAPredicted

34.6 / 100

Open LLM: IFEvalPredicted

76.3 / 100

Open LLM: MATH Level 5Predicted

36.6 / 100

Open LLM: MMLU-PROPredicted

50.8 / 100

Open LLM: MUSRPredicted

44.8 / 100

RE-BenchPredicted

0.4

SimpleBenchPredicted

25.4 / 100

SWE-bench LitePredicted

12.6 / 100

tau-bench RetailPredicted

74.2 / 100

Vectara Factual ConsistencyPredicted

93.7 / 100

Vectara Hallucination RatePredicted

6.3 / 100

WebArenaPredicted

18.6 / 100

WildBenchPredicted

53.6

Arena ELO: Codingchatbot_arena

117.0

Arena ELO: Creative Writingchatbot_arena

1389.0

Chatbot Arena ELOchatbot_arena

1396.0

Aider Polyglotepoch_ai

0.0

FictionLiveBenchepoch_ai

0.6

Lech Mazur Writingepoch_ai

7.7

METR Time Horizonsepoch_ai

0.5

OTIS Mock AIME 2024–2025epoch_ai

0.4

HuggingFace Downloads (30d)hf-downloads

788009.0

HuggingFace Likeshf-downloads

3135.0

SWE-bench Verifiedswebench

42.0 / 100

WeirdMLweirdml

36.1 / 100