What are the hardware requirements for Llama 3.2 3B Instruct?

Llama 3.2 3B Instruct requires approximately 6.5 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 3.2 3B Instruct and how large is it?

Llama 3.2 3B Instruct is developed by meta-llama, a model with 3b 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 3.2 3B Instruct?

Llama 3.2 3B Instruct is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Llama 3.2 3B Instruct by meta-llama — VRAM 6.5GB, 128K context

Name: Llama 3.2 3B Instruct
Rating: 1 (1 reviews)
Author: meta-llama

Llama 3.2 3B Instruct is an open-source language model by meta-llama. Features: 3b LLM, VRAM: 6.5GB, Context: 128K, License: llama3.2, Instruction-Based, LLM Explorer Score: 0.4.

Model Card on HF 🤗: https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct

Llama 3.2 3B Instruct Benchmarks

IFEval: 73.93 vs 88 (so35)^-16%

MATH Lvl 5: 17.67

LLME Score: 0.39756

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

Llama 3.2 3B Instruct (meta-llama/Llama-3.2-3B-Instruct)

🌟 Advertise your project 🚀

Llama 3.2 3B Instruct Parameters and Internals

Model Type

text-generation

Use Cases

Applications:

multilingual dialogue, knowledge retrieval, summarization, AI powered writing assistants, query and prompt rewriting

Limitations:

Use in any manner that violates applicable laws or regulations

Supported Languages

English (officially supported), German (officially supported), French (officially supported), Italian (officially supported), Portuguese (officially supported), Hindi (officially supported), Spanish (officially supported), Thai (officially supported)

Training Details

Data Sources:

publicly available online data

Data Volume:

up to 9 trillion tokens

Methodology:

Supervised Fine-Tuning (SFT), Rejection Sampling (RS), and Direct Preference Optimization (DPO)

Context Length:

128000

Hardware Used:

H100-80GB GPUs (TDP of 700W)

Model Architecture:

Auto-regressive transformer with Grouped-Query Attention (GQA)

Input Output

Accepted Modalities:

Multilingual Text

Output Format:

Multilingual Text and code

LLM Name	Llama 3.2 3B Instruct
Repository 🤗	https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct
Model Size	3b
Required VRAM	6.5 GB
Updated	2025-09-23
Maintainer	meta-llama
Model Type	llama
Instruction-Based	Yes
Model Files	5.0 GB: 1-of-2 1.5 GB: 2-of-2
Supported Languages	en de fr it pt hi es th
Model Architecture	LlamaForCausalLM
License	llama3.2
Context Length	131072
Model Max Length	131072
Transformers Version	4.45.0.dev0
Tokenizer Class	PreTrainedTokenizerFast
Vocabulary Size	128256
Torch Data Type	bfloat16

Quantized Models of the Llama 3.2 3B Instruct

Model	Likes	Downloads	VRAM
...2 3B Instruct Unsloth Bnb 4bit	10	106422	2 GB
Gladiator Mini Exp 1211 3B	0	14	6 GB
Orca Mini V9 0 3B Instruct	5	17	6 GB
...r Mini Exp 1221 3B Instruct V2	0	13	6 GB
...ator Mini Exp 1222 3B Instruct	0	12	6 GB
Llama 3.2 3B Instruct GGUF	62	78782	0 GB
...ator Mini Exp 1221 3B Instruct	0	16	6 GB
Llama 3.2 MEDIT 3B O1	13	185	7 GB
Llama 3.2 3B Instruct GGUF	15	62199	0 GB
Komodo Llama 3.2 3B V2 Fp16	6	19	6 GB

Best Alternatives to Llama 3.2 3B Instruct

Best Alternatives	Context / RAM	Downloads	Likes
Llama 3.2 3B Instruct	128K / 6.5 GB	1831	9
Llama 3.2 3B RP Toxic Fuse	128K / 6.4 GB	14	2
DeepSeek R1 Distill Llama 3B	128K / 6.5 GB	194	16
Llama 3.2 3B Instruct	128K / 6.5 GB	218675	88
Llama Email Fraud Detector	128K / 6.4 GB	8	0
FuseChat Llama 3.2 3B Instruct	128K / 6.5 GB	29	7
ReasoningCore 3B T1 1	128K / 6.5 GB	23	1
... 3.2 3B Math Instruct RE1 ORPO	128K / 6.5 GB	48	0
Zeitgeist 3B V1	128K / 6.5 GB	105	5
Llama 3.2 3B ToxicKod	128K / 6.4 GB	16	2

Rank the Llama 3.2 3B Instruct Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

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

What open-source LLMs or SLMs are you in search of? 53444 in total.

Email us: info@extractum.io. Our Privacy Policy | Terms and Conditions | Suggest an improvement.

Our Social Media →

Original data from HuggingFace, OpenCompass and various public git repos.

Check out Ag3ntum — our secure, self-hosted AI agent for server management.

Release v20260328a

Support LLM Explorer

Llama 3.2 3B Instruct by meta-llama