Project AURIONAI Use-Cases Research & Innovation
LLM

Llama-3.1-Nemotron-70B-Instruct

NVIDIA customization of Llama 3.1 70B Instruct trained with RLHF (REINFORCE) to improve the helpfulness of responses. Use is governed by the NVIDIA Open Model License with the Llama 3.1 Community License as additional terms.

NVIDIANVIDIAUSANemotronFamily
NVIDIA
Levelmodel
ClassLLM
Parametersnot in the notes
Context128,000
Released12 Oct 2024
Opennessopen weights restricted
Runs onmulti-GPU node
Statusactive

Benchmark Results

vendor-reported
Arena Text leaderboardhuman preference arena
Chatbot Arena score quoted by NVIDIA on the model card; style controlled rank 26vendor-reported
1,267 +-7
1,2001,4001,600

Score in Elo; the axis starts at 1,200, as Elo has no zero. Higher is better.

rank 9, Chatbot Arena leaderboard as of 24 Oct 202424 Oct 2024 huggingface.co
Where this comes from
Every score names who ran it: vendor-reported, maintainer-run or independent. A rank is shown with the population it was ranked in (gap rule 3).

Lineage

continued pretrainfinetune
Llama 3.1 70B Instruct to Llama-Sahabat-AI-v2-70B-IT: continued pretrain, medium confidenceLlama 3.1 70B Instruct to Llama-3.1-Nemotron-70B-Instruct: finetune, high confidenceLlama 3.1 70B InstructLlama 3.1 70B InstructMetaLlama-3.1-Nemotron-70B-InstructLlama-3.1-Nemotron-70B-Ins…NVIDIA, Oct 2024Llama-Sahabat-AI-v2-70B-ITLlama-Sahabat-AI-v2-70B-ITGoTo Gojek Tokopedia, Jun 2025
MetaLlama 3.1 70B Instruct is the parent (finetune)high confidenceHugging Face base_model
Where this comes from
Parent and child relations from model cards, Hugging Face base_model fields and the genealogy tables of the research notes. The tree shows two generations each way, 3 models. Confidence and evidence are the notes' own.

R&D Topics

Open-weight frontier releases as research substrateused by

Runtimes

NVIDIA NIMnvidia

Identity

Developer
NVIDIA
Family
Nemotron
Released
12 Oct 2024 sourcesregistry date
Weights published
12 Oct 2024
Record checked
11 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture
not in the notes
Total parameters
– sources
Active parameters
–
Checkpoint parameters
–
Note
70B per model name
Context length
128,000 tokens
Input
text
Output
text
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness
open weights restricted
Licence
NVIDIA Open Model License sources
Access routes
open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
multi-GPU node
Footprint
Needs 2 or more 80GB GPUs (Ampere or newer) and 150GB disk with Transformers; also offered as an NVIDIA NIM container
Runtimes
NVIDIA NIM
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support
not stated
Regional languages
not in the notes
Jakarta hosting
unknown
Languages
Hugging Face language tag: en
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.

Industry Use Cases

No industry use case links this model.

Sources

  1. huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Instruct-HF https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
  2. catalog.ngc.nvidia.com/orgs/nim/teams/nvidia/containers/llama-3.1-ne... https://catalog.ngc.nvidia.com/orgs/nim/teams/nvidia/containers/llama-3.1-nemotron-70b-instruct

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Llama-3.1-Nemotron-70B-Instruct", https://aurion.semai.id/models/nvidia-llama-3-1-nemotron-70b-instruct/, records of October 2026.