Frontier Log

NVIDIAModel

Llama Nemotron Super 49B v1

Made by
NVIDIA
Released
Announced in
NVIDIA Launches Family of Open Reasoning AI Models for Developers and Enterprises to Build Agentic AI Platformsnvidianews.nvidia.com
Event
Llama Nemotron
API record
/api/v1/models/llama-nemotron-super-49b-v1

Benchmark results

BenchmarkScoreSource check
GPQA Diamondreasoning on; average of up to 16 runs; 32K sequence length66.67%
CorroboratedThe same number is printed in 1 other lab document: 1
reported later, in Build More Accurate and Efficient AI Agents with the New NVIDIA Llama Nemotron Super v1.566.7%
CorroboratedThe same number is printed in 3 other lab documents: 1 2 3
BF16 baseline of an NVFP4 quantization study; eval tools differ from the model card; average of 20 runs · reported later, in Quantization-Aware Distillation for NVFP4 Inference Accuracy Recovery66.5%
CheckedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced.
AIME 2024No tools · reported later, in Build More Accurate and Efficient AI Agents with the New NVIDIA Llama Nemotron Super v1.567.5%
CorroboratedThe same number is printed in 3 other lab documents: 1 2 3
AIME 2025No tools · reasoning on; average of up to 16 runs; 32K sequence length58.4%
CheckedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced.
reported later, in Build More Accurate and Efficient AI Agents with the New NVIDIA Llama Nemotron Super v1.553.3%
CorroboratedThe same number is printed in 2 other lab documents: 1 2
BF16 baseline of an NVFP4 quantization study; eval tools differ from the model card; average of 48 runs · reported later, in Quantization-Aware Distillation for NVFP4 Inference Accuracy Recovery46.0%
CheckedA person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced.
LiveCodeBenchproblem window not stated · reported later, in Build More Accurate and Efficient AI Agents with the New NVIDIA Llama Nemotron Super v1.545.5%
CorroboratedThe same number is printed in 3 other lab documents: 1 2 3
Sources

Reported by NVIDIA

Events

Stories

All models