Apertus 8B Instruct

September 2, 2025Last synced Apr 7, 2026, 4:00 PM

Blended Price

$0.13/M

Input Price

$0.10/M

Output Price

$0.20/M

Speed

0 tok/s

TTFT

0.00s

Benchmark Scores

Intelligence Index

External Benchmarks

AA-Omniscience Accuracy

How Apertus 8B Instruct 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

0.3 / 100

LiveCodeBench

— / 100

Terminal-Bench Hard

0.0 / 100

τ²-Bench

0.1 / 100

AA-Omniscience Hallucination RatePredicted

97.0 / 100

Aider PolyglotPredicted

18.0 / 100

AIMEPredicted

0.0 / 30

AIME 2025Predicted

0.0 / 30

AlpacaEval 2.0 LCPredicted

17.1 / 100

AlpacaEval 2.0 RawPredicted

13.9 / 100

ARC-AGI-1Predicted

60.3 / 100

ARC-AGI-1 Cost per TaskPredicted

0.5

ARC-AGI-2Predicted

7.1 / 100

ARC-AGI-2 Cost per TaskPredicted

0.6

BFCL (Berkeley Function Calling)Predicted

22.5

BigCodeBench CompletePredicted

25.4 / 100

BigCodeBench InstructPredicted

19.9 / 100

AA Intelligence Index (Matrix)Predicted

49.2

AA Long Context Reasoning (Matrix)Predicted

77.0

AIME 2024Predicted

33.4

AIME 2025 (Matrix)Predicted

84.7

Arena-Hard AutoPredicted

14.5

BrowseCompPredicted

87.8

BRUMO 2025Predicted

99.7

CMIMC 2025Predicted

94.1

CritPtPredicted

12.5

GPQA Diamond (Matrix)Predicted

45.9

HLE (Matrix)Predicted

34.0

HMMT Feb 2025Predicted

71.9

HMMT Nov 2025Predicted

94.3

HumanEvalPredicted

61.4

IFBench (Matrix)Predicted

37.1

IFEvalPredicted

74.7

IMO 2025Predicted

45.3

LiveCodeBench (Matrix)Predicted

33.3

MATH-500 (Matrix)Predicted

85.0

MathArena Apex 2025Predicted

19.0

MMLU-Pro (Matrix)Predicted

43.3

MMMU-ProPredicted

75.5

MRCR v2Predicted

81.7

OSWorldPredicted

80.1

SimpleQAPredicted

49.7

SMT 2025Predicted

93.3

SWE-bench ProPredicted

58.0

Tau-Bench Telecom (Matrix)Predicted

99.0

Terminal-Bench 2.0Predicted

76.4

Terminal-Bench 1.0Predicted

31.3

USAMO 2025Predicted

11.0

Video-MMUPredicted

83.8

browsecompPredicted

89.6

BullshitBenchPredicted

44.7 / 100

Aider PolyglotPredicted

0.0

Apex AgentsPredicted

4.5

Arc Agi 2Predicted

0.2

BIG-Bench HardPredicted

3.0

CAD-EvalPredicted

8.1

Chess PuzzlesPredicted

0.4

CyBenchPredicted

0.2

DeepResearchBenchPredicted

0.4

FictionLiveBenchPredicted

0.4

GeoBenchPredicted

0.0

GSM8K (Epoch)Predicted

7.1

HellaSwagPredicted

0.0

Lech Mazur WritingPredicted

7.0

METR Time HorizonsPredicted

10.1

OTIS Mock AIME 2024–2025Predicted

0.0

PosttrainbenchPredicted

0.0

SimpleQA Verified (Epoch)Predicted

0.6

The Agent CompanyPredicted

0.9

TriviaQAPredicted

10.3

WinoGrandePredicted

0.7

FrontierMathPredicted

43.3 / 100

GAIA Level 1Predicted

6.5

GAIA Level 2Predicted

0.4

GAIA Level 3Predicted

0.0

GAIAPredicted

3.8 / 100

LegalBenchPredicted

20.7 / 100

LiveBench CodingPredicted

78.3 / 100

LiveBench Data AnalysisPredicted

69.7 / 100

LiveBench LanguagePredicted

81.1 / 100

LiveBench MathPredicted

86.6 / 100

LiveBench OverallPredicted

74.0 / 100

LiveBench ReasoningPredicted

78.3 / 100

LiveCodeBenchPredicted

0.1 / 100

LongBench v2 EasyPredicted

28.2

LongBench v2 HardPredicted

23.6

LongBench v2Predicted

22.1 / 100

MATH-500Predicted

0.2 / 100

MathVistaPredicted

18.2 / 100

MedQA (USMLE)Predicted

57.7

MLE-benchPredicted

62.1 / 100

MMLU ProPredicted

0.4 / 100

MMMUPredicted

54.1 / 100

MMTU Table UnderstandingPredicted

49.1 / 100

MT-BenchPredicted

7.6 / 10

NoLiMa (NIAH)Predicted

80.7 / 100

OCRBench v2Predicted

53.1 / 100

Open LLM AveragePredicted

17.3 / 100

Open LLM: BBHPredicted

40.7 / 100

Open LLM: GPQAPredicted

26.8 / 100

Open LLM: IFEvalPredicted

57.9 / 100

Open LLM: MATH Level 5Predicted

7.3 / 100

Open LLM: MMLU-PROPredicted

24.9 / 100

Open LLM: MUSRPredicted

36.8 / 100

RE-BenchPredicted

100.0

SimpleBenchPredicted

42.7 / 100

simpleqaPredicted

37.3

SWE-bench LitePredicted

8.6 / 100

SWE-bench VerifiedPredicted

46.4 / 100

tau-bench RetailPredicted

92.2 / 100

Vectara Factual ConsistencyPredicted

87.2 / 100

Vectara Hallucination RatePredicted

12.8 / 100

WebArenaPredicted

0.0 / 100

WeirdMLPredicted

19.5 / 100

WildBenchPredicted

26.9