Reka Flash (Sep '24)

October 4, 2024Last synced Jun 12, 2026, 8:42 PM

Human Frontier

72.2

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

Blended Price

$0.35/M

Input Price

$0.20/M

Output Price

$0.80/M

Speed

85 tok/s

TTFT

1.76s

Benchmark Scores

Intelligence Index

External Benchmarks

AA-Omniscience Accuracy

How Reka Flash (Sep '24) Compares

Axes

X Axis

Y Axis

Bubble Size

Filters

Blended Price (USD)$0.00 – $37.5

AgMoBench Overall≥ 2.0

Providers

ai2ai21-labsalibabaanthropicarceeawsazurebaidu

Show quadrants

alibaba

mistral

anthropic

openai

meta

liquidai

nvidia

azure

google

deepseek

stepfun

cohere

minimax

ibm

nous-research

multiversecomputing

kimi

aws

zai

deepcogito

kwaikat

xai

xiaomi

ai21-labs

inclusionai

baidu

sarvam

prime-intellect

bytedance_seed

inception

upstage

reka-ai

tencent

arcee

ai2

swiss-ai-initiative

perplexity

Bubble size = Context Window

8191.00

2000000.00

Compare with other models

GPQA Diamond

— / 100

LiveCodeBench

— / 100

Terminal-Bench Hard

— / 100

τ²-Bench

— / 100

AA-Omniscience Hallucination RatePredicted

96.1 / 100

Aider PolyglotPredicted

18.2 / 100

AIMEPredicted

0.0 / 30

AIME 2025Predicted

0.0 / 30

AlpacaEval 2.0 LCPredicted

24.2 / 100

AlpacaEval 2.0 RawPredicted

19.2 / 100

ARC-AGI-1Predicted

22.9 / 100

ARC-AGI-1 Cost per TaskPredicted

0.1

ARC-AGI-2Predicted

60.1 / 100

ARC-AGI-2 Cost per TaskPredicted

0.1

BFCL (Berkeley Function Calling)Predicted

26.7

BigCodeBench CompletePredicted

36.5 / 100

BigCodeBench InstructPredicted

28.9 / 100

AA Intelligence Index (Matrix)Predicted

56.2

AA Long Context Reasoning (Matrix)Predicted

74.3

AIME 2024Predicted

28.8

AIME 2025 (Matrix)Predicted

64.3

Arena-Hard AutoPredicted

27.0

BrowseCompPredicted

78.7

BRUMO 2025Predicted

97.8

CMIMC 2025Predicted

87.2

GPQA Diamond (Matrix)Predicted

48.9

HLE (Matrix)Predicted

22.3

HMMT Feb 2025Predicted

56.8

HMMT Nov 2025Predicted

92.3

HumanEvalPredicted

74.5

IFBench (Matrix)Predicted

34.0

IFEvalPredicted

76.8

IMO 2025Predicted

26.9

LiveCodeBench (Matrix)Predicted

37.1

MATH-500 (Matrix)Predicted

87.9

MathArena Apex 2025Predicted

5.1

MMLU-Pro (Matrix)Predicted

61.6

MMMU-ProPredicted

77.4

MRCR v2Predicted

78.8

OSWorldPredicted

66.1

SimpleQAPredicted

35.5

SMT 2025Predicted

88.3

SWE-bench ProPredicted

48.7

Tau-Bench Telecom (Matrix)Predicted

98.1

Terminal-Bench 2.0Predicted

51.3

Terminal-Bench 1.0Predicted

22.2

USAMO 2025Predicted

7.9

Video-MMUPredicted

84.7

browsecompPredicted

79.7

BullshitBenchPredicted

27.6 / 100

Aider PolyglotPredicted

0.0

Apex AgentsPredicted

3.4

Arc Agi 2Predicted

20.5

BIG-Bench HardPredicted

3.0

CAD-EvalPredicted

7.6

Chess PuzzlesPredicted

0.2

CyBenchPredicted

0.1

DeepResearchBenchPredicted

0.4

FictionLiveBenchPredicted

0.4

GeoBenchPredicted

0.0

GSM8K (Epoch)Predicted

0.1

HellaSwagPredicted

0.3

Lech Mazur WritingPredicted

7.0

METR Time HorizonsPredicted

2.3

OTIS Mock AIME 2024–2025Predicted

0.0

PosttrainbenchPredicted

0.0

SimpleQA Verified (Epoch)Predicted

0.4

The Agent CompanyPredicted

1.0

TriviaQAPredicted

5.8

WinoGrandePredicted

0.8

FrontierMathPredicted

26.7 / 100

GAIA Level 1Predicted

12.4

GAIA Level 2Predicted

1.6

GAIA Level 3Predicted

0.0

GAIAPredicted

7.4 / 100

GPQA DiamondPredicted

0.3 / 100

HLEPredicted

0.0 / 100

IFBenchPredicted

0.3 / 100

LCRPredicted

0.0 / 100

LegalBenchPredicted

45.5 / 100

LiveBench CodingPredicted

72.9 / 100

LiveBench Data AnalysisPredicted

57.4 / 100

LiveBench LanguagePredicted

69.3 / 100

LiveBench MathPredicted

76.8 / 100

LiveBench OverallPredicted

61.3 / 100

LiveBench ReasoningPredicted

61.3 / 100

LiveCodeBenchPredicted

0.1 / 100

LongBench v2 EasyPredicted

33.2

LongBench v2 HardPredicted

29.5

LongBench v2Predicted

27.1 / 100

MathVistaPredicted

47.8 / 100

MedQA (USMLE)Predicted

63.7

MLE-benchPredicted

39.8 / 100

MMLU ProPredicted

0.4 / 100

MMTU Table UnderstandingPredicted

48.4 / 100

MT-BenchPredicted

7.9 / 10

NoLiMa (NIAH)Predicted

86.2 / 100

OCRBench v2Predicted

64.5 / 100

Open LLM AveragePredicted

25.6 / 100

Open LLM: BBHPredicted

51.5 / 100

Open LLM: GPQAPredicted

29.9 / 100

Open LLM: IFEvalPredicted

65.3 / 100

Open LLM: MATH Level 5Predicted

16.4 / 100

Open LLM: MMLU-PROPredicted

35.0 / 100

Open LLM: MUSRPredicted

39.9 / 100

RE-BenchPredicted

89.9

SciCodePredicted

0.1 / 100

SimpleBenchPredicted

32.2 / 100

simpleqaPredicted

28.5

SWE-bench LitePredicted

10.0 / 100

SWE-bench VerifiedPredicted

40.1 / 100

τ²-BenchPredicted

0.0 / 100

tau-bench RetailPredicted

87.1 / 100

Terminal-Bench HardPredicted

0.0 / 100

Vectara Factual ConsistencyPredicted

92.1 / 100

Vectara Hallucination RatePredicted

7.9 / 100

WebArenaPredicted

0.3 / 100

WeirdMLPredicted

19.4 / 100

Arena ELO: Creative Writingchatbot_arena

1242.0

Chatbot Arena ELOchatbot_arena

1272.0

WildBenchwildbench

34.6