Qwen3 14B (Reasoning)

April 28, 2025Last synced Aug 1, 2026, 12:19 AM

Human Frontier

—

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

Blended Price

$1.31/M

Input Price

$0.35/M

Output Price

$4.20/M

Speed

—

TTFT

—

Benchmark Scores

Intelligence Index

Performance Radar

Qwen3 14B (Reasoning)

Field Average

External Benchmarks

AIME 2024benchmark_matrix

How Qwen3 14B (Reasoning) Compares

Axes

X Axis

Y Axis

Bubble Size

Filters

Blended Price (USD)$0.00 – $37.5

AgMoBench Overall≥ 1.8

Providers

ai2ai21-labsalibabaanthropicarceeawsazurebaidu

Show quadrants

anthropic

openai

mistral

azure

cohere

meta

ai21-labs

inclusionai

google

ai2

ibm

alibaba

upstage

nvidia

aws

kimi

nous-research

thinking-machines

minimax

stepfun

tencent

deepseek

zai

multiversecomputing

xiaomi

xai

deepcogito

sarvam

kwaikat

baidu

swiss-ai-initiative

bytedance_seed

inception

reka-ai

arcee

perplexity

Bubble size = Context Window

8191.00

2000000.00

Compare with other models

GPQA Diamond

60.4 / 100

LiveCodeBench

52.3 / 100

Terminal-Bench Hard

3.8 / 100

τ²-Bench

34.5 / 100

AIME 2025 (Matrix)benchmark_matrix

72.0

Arena-Hard Autobenchmark_matrix

85.5

Codeforces Ratingbenchmark_matrix

1900.0

GPQA Diamond (Matrix)benchmark_matrix

64.0

GSM8Kbenchmark_matrix

92.5

HumanEvalbenchmark_matrix

85.0

LiveCodeBench (Matrix)benchmark_matrix

55.0

MATH-500 (Matrix)benchmark_matrix

95.0

MMLUbenchmark_matrix

81.0

MMLU-Pro (Matrix)benchmark_matrix

61.0

AA-Omniscience AccuracyPredicted

43.1 / 100

AA-Omniscience Hallucination RatePredicted

94.3 / 100

Aider PolyglotPredicted

45.6 / 100

AlpacaEval 2.0 LCPredicted

56.8 / 100

AlpacaEval 2.0 RawPredicted

49.2 / 100

ARC-AGI-1Predicted

24.8 / 100

ARC-AGI-1 Cost per TaskPredicted

0.1

ARC-AGI-2Predicted

0.4 / 100

ARC-AGI-2 Cost per TaskPredicted

0.2

BigCodeBench CompletePredicted

47.9 / 100

BigCodeBench InstructPredicted

38.4 / 100

AA Intelligence Index (Matrix)Predicted

63.1

AA Long Context Reasoning (Matrix)Predicted

73.0

BrowseCompPredicted

69.1

BRUMO 2025Predicted

95.2

CMIMC 2025Predicted

88.6

HLE (Matrix)Predicted

21.2

HMMT Feb 2025Predicted

70.8

HMMT Nov 2025Predicted

91.1

IFBench (Matrix)Predicted

37.2

IFEvalPredicted

84.5

IMO 2025Predicted

22.4

MathArena Apex 2025Predicted

2.1

MMMU-ProPredicted

79.9

MRCR v2Predicted

77.5

OSWorld (Unversioned Matrix Import)Predicted

41.9 / 100

SimpleQAPredicted

33.9

SMT 2025Predicted

86.8

SWE-bench ProPredicted

38.1

Tau-Bench Telecom (Matrix)Predicted

97.2

Terminal-Bench 2.0Predicted

35.1

Terminal-Bench 1.0Predicted

23.7

USAMO 2025Predicted

11.1

Video-MMUPredicted

83.1

browsecompPredicted

67.8

BullshitBenchPredicted

50.6 / 100

Aider PolyglotPredicted

0.0

Apex AgentsPredicted

2.6

Arc Agi 2Predicted

0.3

BIG-Bench HardPredicted

3.0

CAD-EvalPredicted

4.8

Chess PuzzlesPredicted

0.2

CyBenchPredicted

0.2

DeepResearchBenchPredicted

0.4

FictionLiveBenchPredicted

0.7

GeoBenchPredicted

0.0

GSM8K (Epoch)Predicted

5.1

HellaSwagPredicted

0.1

Lech Mazur WritingPredicted

7.6

METR Time HorizonsPredicted

2.2

OTIS Mock AIME 2024–2025Predicted

0.2

PosttrainbenchPredicted

0.0

SimpleQA Verified (Epoch)Predicted

0.3

The Agent CompanyPredicted

1.7

TriviaQAPredicted

19.4

WinoGrandePredicted

0.9

FrontierMathPredicted

31.0 / 100

GAIA Level 1Predicted

46.6

GAIA Level 2Predicted

20.9

GAIA Level 3Predicted

1.5

GAIAPredicted

31.1 / 100

LegalBenchPredicted

76.2 / 100

LiveBench CodingPredicted

70.0 / 100

LiveBench Data AnalysisPredicted

51.7 / 100

LiveBench LanguagePredicted

63.0 / 100

LiveBench MathPredicted

70.9 / 100

LiveBench OverallPredicted

55.5 / 100

LiveBench ReasoningPredicted

53.8 / 100

LongBench v2 EasyPredicted

45.7

LongBench v2 HardPredicted

37.4

LongBench v2Predicted

31.3 / 100

MathVistaPredicted

57.7 / 100

MedQA (USMLE)Predicted

85.0

MLE-bench (Original)Predicted

33.8 / 100

MMMUPredicted

69.7 / 100

MMTU Table UnderstandingPredicted

56.9 / 100

MT-BenchPredicted

7.9 / 10

NoLiMa (NIAH)Predicted

93.4 / 100

OCRBench v2Predicted

77.7 / 100

RE-BenchPredicted

57.0

SimpleBenchPredicted

37.7 / 100

simpleqaPredicted

34.9

SWE-bench LitePredicted

24.8 / 100

SWE-bench VerifiedPredicted

54.4 / 100

tau-bench RetailPredicted

83.3 / 100

WebArenaPredicted

1.5 / 100

WeirdMLPredicted

31.6 / 100

WildBenchPredicted

43.0

BFCL (Berkeley Function Calling)bfcl

41.0

Epoch Capabilities Indexepoch_ai

3.1968e+24

HuggingFace Downloads (30d)hf-downloads

2951714.0

HuggingFace Likeshf-downloads

436.0

Open LLM Averageopen_llm_leaderboard

17.2 / 100

Open LLM: BBHopen_llm_leaderboard

47.0 / 100

Open LLM: GPQAopen_llm_leaderboard

30.3 / 100

Open LLM: IFEvalopen_llm_leaderboard

27.9 / 100

Open LLM: MATH Level 5open_llm_leaderboard

7.6 / 100

Open LLM: MMLU-PROopen_llm_leaderboard

32.2 / 100

Open LLM: MUSRopen_llm_leaderboard

41.5 / 100

Vectara Factual Consistencyvectara_hallucination

94.6 / 100

Vectara Hallucination Ratevectara_hallucination

5.4 / 100