What are the hardware requirements for Llama 3 3 Nemotron Super 49B V1.5 FP8?

Llama 3 3 Nemotron Super 49B V1.5 FP8 requires approximately 52 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Llama 3 3 Nemotron Super 49B V1.5 FP8 and how large is it?

Llama 3 3 Nemotron Super 49B V1.5 FP8 is developed by nvidia, a model with 49b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Llama 3 3 Nemotron Super 49B V1.5 FP8?

Llama 3 3 Nemotron Super 49B V1.5 FP8 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Llama 3 3 Nemotron Super 49B V1.5 FP8 by nvidia — VRAM 52GB, 128K context

Name: Llama 3 3 Nemotron Super 49B V1.5 FP8
Author: nvidia

Llama 3 3 Nemotron Super 49B V1.5 FP8 is an open-source language model by nvidia. Features: 49b LLM, VRAM: 52GB, Context: 128K, License: other, LLM Explorer Score: 0.28.

Arxiv:2411.19146 Arxiv:2502.00203 Arxiv:2505.00949 Conversational Custom code En Llama-3 Nemotron-nas Nvidia Pytorch Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/nvidia/Llama-3_3-Nemotron-Super-49B-v1_5-FP8

Llama 3 3 Nemotron Super 49B V1.5 FP8 Parameters and Internals

LLM Name	Llama 3 3 Nemotron Super 49B V1.5 FP8
Repository 🤗	https://huggingface.co/nvidia/Llama-3_3-Nemotron-Super-49B-v1_5-FP8
Model Size	49b
Required VRAM	52 GB
Updated	2026-08-04
Maintainer	nvidia
Model Type	nemotron-nas
Model Files	4.9 GB: 1-of-11 5.0 GB: 2-of-11 5.0 GB: 3-of-11 4.9 GB: 4-of-11 4.9 GB: 5-of-11 4.9 GB: 6-of-11 5.0 GB: 7-of-11 4.9 GB: 8-of-11 4.8 GB: 9-of-11 4.9 GB: 10-of-11 2.8 GB: 11-of-11
Supported Languages	en
Model Architecture	DeciLMForCausalLM
License	other
Context Length	131072
Model Max Length	131072
Transformers Version	4.48.3
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|eot_id\|>
Vocabulary Size	128256
Torch Data Type	bfloat16

Rank the Llama 3 3 Nemotron Super 49B V1.5 FP8 capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Llama 3 3 Nemotron Super 49B V1.5 FP8

Best Alternatives	Context / RAM	Downloads	Likes
... Nemotron Super 49B V1.5 NVFP4	128K / 31.1 GB	22695	20
...lama 3 3 Nemotron Super 49B V1	128K / 100.2 GB	189371	325
...ma 3 3 Nemotron Super 49B V1.5	128K / 100.2 GB	21	3
...lama 3 3 Nemotron Super 49B V1	128K / 100.2 GB	50	0
...ma 3 3 Nemotron Super 49B V1.5	128K / 100.2 GB	57679	234
Valkyrie 49B V2.1	128K / 100.2 GB	223	30
Valkyrie 49B V2	128K / 100.2 GB	10	30
... 3 3 Nemotron Super 49B V1 FP8	128K / 52 GB	1025	13
...a 3 3 Nemotron Super 49B GenRM	128K / 99.7 GB	40	19
...n Super 49B GenRM Multilingual	128K / 99.7 GB	81	7

Note: green Score (e.g. "73.2") means that the model is better than nvidia/Llama-3_3-Nemotron-Super-49B-v1_5-FP8.