AI Model
Llama 3.2 3B Instruct
3B parameters · text-generation
VRAM (FP16)
7 GB
VRAM (INT4)
2 GB
Family
llama
Compatible GPUs
NVIDIA T4
Min GPUs: 1 · fp16
NVIDIA P100
Min GPUs: 1 · fp16
NVIDIA RTX 5080
Min GPUs: 1 · fp16
NVIDIA RTX A4000
Min GPUs: 1 · fp16
NVIDIA L4
Min GPUs: 1 · fp16
NVIDIA A10
Min GPUs: 1 · fp16
NVIDIA A10G
Min GPUs: 1 · fp16
NVIDIA RTX 4090
Min GPUs: 1 · fp16
NVIDIA RTX A5000
Min GPUs: 1 · fp16
NVIDIA RTX 3090
Min GPUs: 1 · fp16
NVIDIA V100
Min GPUs: 1 · fp16
NVIDIA RTX 5090
Min GPUs: 1 · fp16
NVIDIA L40S
Min GPUs: 1 · fp16
NVIDIA A40
Min GPUs: 1 · fp16
NVIDIA RTX 6000 Ada
Min GPUs: 1 · fp16
NVIDIA RTX A6000
Min GPUs: 1 · fp16
NVIDIA A16
Min GPUs: 1 · fp16
NVIDIA H100
Min GPUs: 1 · fp16
NVIDIA A100
Min GPUs: 1 · fp16
NVIDIA GH200
Min GPUs: 1 · fp16
AMD Instinct MI250X
Min GPUs: 1 · fp16
NVIDIA H200
Min GPUs: 1 · fp16
NVIDIA H200 NVL
Min GPUs: 1 · fp16
NVIDIA B200
Min GPUs: 1 · fp16
NVIDIA GB200
Min GPUs: 1 · fp16
AMD Instinct MI300X
Min GPUs: 1 · fp16
NVIDIA B300
Min GPUs: 1 · fp16
Supported Frameworks
vLLMPyTorchTensorRT-LLMText Generation InferenceOllamaSGLangllama.cpp
Deploy Llama 3.2 3B Instruct
Get a full deployment stack recommendation — GPU, count, framework, quantization, and projected cost.
Start deploymentHugging Face
View the model card, tokenizer, and weights on the Hugging Face Hub.
Open on Hugging FaceVRAM Usage
FP16 serving needs about 7 GB before workload-specific headroom. INT4 quantization reduces the model weights to about 2 GB, which is the practical path for large models on smaller GPU clusters.
Related Llama 3.2 3B Instruct resources
Move from model requirements into compatible GPU prices, deployment, and the wider model catalog.
NVIDIA T4 prices for Llama 3.2 3B InstructRecommended GPU path for this model at fp16 precision.NVIDIA P100 cloud pricesCompatible option for Llama 3.2 3B Instruct; minimum 1 GPU.NVIDIA RTX 5080 cloud pricesCompatible option for Llama 3.2 3B Instruct; minimum 1 GPU.NVIDIA RTX A4000 cloud pricesCompatible option for Llama 3.2 3B Instruct; minimum 1 GPU.Deploy Llama 3.2 3B InstructGenerate a deployment recommendation with GPU count, framework, and estimated cost.Model VRAM leaderboardCompare FP16 and INT4 memory requirements across other deployable models.