GPT-5 nano (medium)

August 7, 2025Last synced Apr 7, 2026, 4:00 PM

Blended Price

$0.14/M

Input Price

$0.050/M

Output Price

$0.40/M

Speed

148 tok/s

TTFT

49.30s

Benchmark Scores

Intelligence Index

External Benchmarks

ARC-AGI-1

How GPT-5 nano (medium) Compares

Axes

X Axis

Y Axis

Bubble Size

Filters

Blended Price (USD)$0.00 – $30.0

AgMoBench Overall≥ 3.5

Providers

ai21-labsalibabaanthropicawsazurebaiducoheredeepseek

Show quadrants

anthropic

openai

google

meta

mistral

nvidia

kimi

xai

azure

deepseek

aws

cohere

baidu

ai21-labs

zai

reka-ai

xiaomi

alibaba

minimax

ibm

Bubble size = Context Window

16384.00

2000000.00

Compare with other models

GPQA Diamond

0.7 / 100

LiveCodeBench

0.8 / 100

Terminal-Bench Hard

0.2 / 100

τ²-Bench

0.3 / 100

ARC-AGI-1 Cost per Taskarc_agi

0.0

ARC-AGI-2arc_agi

0.0 / 100

ARC-AGI-2 Cost per Taskarc_agi

0.0

AA-Omniscience AccuracyPredicted

36.4 / 100

AA-Omniscience Hallucination RatePredicted

95.9 / 100

Aider PolyglotPredicted

47.6 / 100

AIMEPredicted

0.8 / 30

AlpacaEval 2.0 LCPredicted

16.4 / 100

AlpacaEval 2.0 RawPredicted

12.8 / 100

BigCodeBench CompletePredicted

59.8 / 100

BigCodeBench InstructPredicted

48.7 / 100

AA Intelligence Index (Matrix)Predicted

69.0

AA Long Context Reasoning (Matrix)Predicted

69.6

AIME 2024Predicted

87.1

AIME 2025 (Matrix)Predicted

72.5

Arena-Hard AutoPredicted

81.7

BrowseCompPredicted

50.9

BRUMO 2025Predicted

69.7

CMIMC 2025Predicted

65.1

GPQA Diamond (Matrix)Predicted

75.4

HLE (Matrix)Predicted

12.2

HMMT Feb 2025Predicted

54.4

HMMT Nov 2025Predicted

45.6

HumanEvalPredicted

89.2

IFBench (Matrix)Predicted

44.2

IFEvalPredicted

87.3

IMO 2025Predicted

10.4

LiveCodeBench (Matrix)Predicted

66.9

MATH-500 (Matrix)Predicted

96.6

MathArena Apex 2025Predicted

0.4

MMLU-Pro (Matrix)Predicted

80.9

MMMU-ProPredicted

82.1

MRCR v2Predicted

81.8

OSWorldPredicted

33.8

SimpleQAPredicted

22.1

SMT 2025Predicted

75.6

SWE-bench ProPredicted

32.5

Tau-Bench Telecom (Matrix)Predicted

37.6

Terminal-Bench 2.0Predicted

13.1

Terminal-Bench 1.0Predicted

15.6

USAMO 2025Predicted

7.9

Video-MMUPredicted

86.9

browsecompPredicted

45.9

BullshitBenchPredicted

11.7 / 100

Aider PolyglotPredicted

0.0

Apex AgentsPredicted

1.8

BIG-Bench HardPredicted

3.0

CAD-EvalPredicted

4.4

Chess PuzzlesPredicted

0.1

CyBenchPredicted

0.3

DeepResearchBenchPredicted

0.3

GeoBenchPredicted

0.0

GSM8K (Epoch)Predicted

16.0

HellaSwagPredicted

9.5

Lech Mazur WritingPredicted

7.5

METR Time HorizonsPredicted

0.7

PosttrainbenchPredicted

0.0

The Agent CompanyPredicted

0.9

TriviaQAPredicted

3.2

WinoGrandePredicted

0.9

FrontierMathPredicted

9.4 / 100

GAIA Level 1Predicted

67.8

GAIA Level 2Predicted

56.7

GAIA Level 3Predicted

30.6

GAIAPredicted

51.4 / 100

LegalBenchPredicted

87.3 / 100

LongBench v2 EasyPredicted

58.4

LongBench v2 HardPredicted

49.0

LongBench v2Predicted

44.2 / 100

MATH-500Predicted

1.0 / 100

MathVistaPredicted

67.7 / 100

MedQA (USMLE)Predicted

91.0

MLE-benchPredicted

15.5 / 100

MMMUPredicted

74.8 / 100

MMTU Table UnderstandingPredicted

58.2 / 100

MT-BenchPredicted

8.0 / 10

NoLiMa (NIAH)Predicted

94.8 / 100

OCRBench v2Predicted

84.3 / 100

Open LLM AveragePredicted

38.6 / 100

Open LLM: BBHPredicted

66.6 / 100

Open LLM: GPQAPredicted

35.1 / 100

Open LLM: IFEvalPredicted

64.6 / 100

Open LLM: MATH Level 5Predicted

71.1 / 100

Open LLM: MMLU-PROPredicted

52.2 / 100

Open LLM: MUSRPredicted

45.4 / 100

RE-BenchPredicted

0.2

SimpleBenchPredicted

33.1 / 100

simpleqaPredicted

23.7

SWE-bench LitePredicted

29.4 / 100

tau-bench RetailPredicted

72.3 / 100

WebArenaPredicted

13.8 / 100

WildBenchPredicted

56.2

BFCL (Berkeley Function Calling)bfcl

51.5

Arc Agi 2epoch_ai

0.0

Epoch Capabilities Indexepoch_ai

140.4

FictionLiveBenchepoch_ai

0.4

OTIS Mock AIME 2024–2025epoch_ai

0.7

SimpleQA Verified (Epoch)epoch_ai

0.1

LiveBench Codinglivebench

62.4 / 100

LiveBench Data Analysislivebench

43.4 / 100

LiveBench Languagelivebench

46.8 / 100

LiveBench Mathlivebench

68.4 / 100

LiveBench Overalllivebench

48.6 / 100

LiveBench Reasoninglivebench

40.3 / 100

OpenRouter Token Volumeopenrouter_rankings

287000000000.0

OpenRouter Weekly Trendopenrouter_rankings

10.0

SWE-bench Verifiedswebench

34.8 / 100

Vectara Factual Consistencyvectara_hallucination

89.5 / 100

Vectara Hallucination Ratevectara_hallucination

10.5 / 100

WeirdMLweirdml

38.1 / 100