Released date

Model

Last Updated 3/18/2025

together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561

Llama 3.3 Nemotron Super (Nonthinking)

Nvidia's 49B Nemotron model, fine-tuned with Llama 3.3 weights.

Released Date: 3/18/2025

Avg. Accuracy:

54.1%

Latency:

16.90s

Performance by Benchmark

Benchmarks

Accuracy

Rankings

CaseLaw

69.2%

( 54 / 66 )

69.2%

54 / 66

ContractLaw

73.3%

( 3 / 72 )

73.3%

3 / 72

TaxEval

65.2%

( 40 / 53 )

65.2%

40 / 53

Math500

71.2%

( 41 / 49 )

71.2%

41 / 49

AIME

9.4%

( 37 / 43 )

9.4%

37 / 43

GPQA

40.9%

( 37 / 44 )

40.9%

37 / 44

MMLU Pro

67.0%

( 33 / 43 )

67.0%

33 / 43

LiveCodeBench

36.3%

( 38 / 45 )

36.3%

38 / 45

Academic Benchmarks

Proprietary Benchmarks (contact us to get access)

Cost Analysis

Cost information not available for this model

Performance by Benchmark

Cost Analysis

Join our mailing list to receive benchmark updates on