Llama-3.3-Nemotron-Super-49B-v1 is a model which offers a great tradeoff between model accuracy and efficiency. Efficiency (throughput) directly translates to savings. Using a novel Neural Architecture Search (NAS) approach, we greatly reduce the model's memory footprint, enabling larger workloads, as well as fitting the model on a single GPU at high workloads (H200). This NAS approach enables the selection of a desired point in the accuracy-efficiency tradeoff. For more information on the NAS approach, please refer to this paper.
Added Aug 8, 2025
Context Window
128.0K
Max Output
16.4K
Input Price (Auto)
$0.15/1M
Output Price (Auto)
$0.15/1M
Performance metrics and benchmarks
Sourced from Artificial Analysis.
Intelligence Index
14.3
Auto routing is available for this model. Explicit provider selection is not available.
Loading provider options…
Coding Index
7.6
GPQA Diamond
Graduate-level scientific reasoning
51.7%
Better than 36% of models compared
HLE
Humanity's Last Exam
3.5%
Better than 5% of models compared
IFBench
Instruction-following benchmark
39.5%
Better than 45% of models compared
AA-LCR
Long context reasoning evaluation
11.3%
Better than 26% of models compared
CritPt
Research-level physics reasoning
0.0%
Better than 36% of models compared
SciCode
Python programming for scientific computing
22.9%
Better than 31% of models compared
Terminal-Bench Hard
Agentic coding and terminal use
0.0%
Better than 6% of models compared
AIME 2025
American Invitational Mathematics Examination 2025
7.7%
Better than 12% of models compared
AIME
American Invitational Mathematics Examination
19.3%
Better than 46% of models compared
MMLU-Pro
Professional and academic subject knowledge
69.8%
Better than 36% of models compared
AA-Omniscience Accuracy
Proportion of correctly answered questions
13.5%
Better than 25% of models compared
Last updated May 15, 2026
Artificial AnalysisLiveCodeBench
Contamination-free coding benchmark
28.0%
Better than 31% of models compared
Math-500
Diverse mathematical problem solving benchmark
77.5%
Better than 42% of models compared
AA-Omniscience Hallucination Rate
Rate of incorrect answers among non-correct responses
73.0%
Better than 75% of models compared
