Meta/Dense

Llama-3.2-90B-Vision-Instruct

Name: Llama-3.2-90B-Vision-Instruct
Author: Meta

Llama 3.2 90B Vision Instruct — top open multimodal model for complex visual tasks.

visionchatThinkingDistilled

90B

Parameters

128K

Context length

Benchmarks

Quantizations

450K

HF downloads

Architecture

Dense

Released

2024-09-25

Layers

KV Heads

Head Dim

128

Family

llama

Quantization Options

Quant	Bits	VRAM	Quality
IQ2_M	2.93	33.5 GB	low
Q2_K	3.16	36.0 GB	low
IQ3_XXS	3.25	37.1 GB	low
IQ3_XS	3.5	39.9 GB	low
Q3_K_S	3.64	41.4 GB	low
IQ3_M	3.76	42.8 GB	low
Q3_K_M	4	45.5 GB	low
Q3_K_L	4.3	48.9 GB	moderate
IQ4_XS	4.46	50.7 GB	moderate
Q4_K_S	4.67	53.0 GB	moderate
Q4_K_M	4.89	55.5 GB	good
Q5_K_S	5.57	63.2 GB	good
Q5_K_M	5.7	64.6 GB	good
Q6_K	6.56	74.3 GB	excellent
Q8_0	8.5	96.1 GB	lossless
FP16	16	180.5 GB	lossless

Select your GPU above to see speed estimates and compatibility for each quantization.

▸ READY TO RUN THIS?RENT BY THE HOUR

RENT A GPU AND RUN LLAMA-3.2-90B-VISION-INSTRUCT NOW

Rent on RunPod →Or Vast.ai →

Spin up an A100 / H100 / 4090 in ~60s. Pay by the second. Cancel anytime.

Community Ratings

Loading ratings...

Benchmarks (13)

MMBench85.5

IFEval82.0

BBH70.0

MMLU-PRO68.0

HumanEval65.0

MMMU60.3

MATH52.0

GPQA48.0

GPQA Diamond43.2

MUSR25.0

LiveCodeBench21.4

AIME5.0

HLE4.9

Run this model

▸Easiest way to get started·Beginners

DOCS ↗

curl -fsSL https://ollama.com/install.sh | sh

$ollama run llama3.2-vision:90b-instruct-q4_K_M

Downloads and runs automatically. Add --verbose for speed stats.

▸ SETUP GUIDE

Auto-setup with fitmyllm CLI

Detects your GPU, recommends the best model, downloads it, and starts chatting — zero config. Benchmarks your speed and contributes anonymous data to improve predictions.

pip install fitmyllmthen run fitmyllmLearn more

Auto-detect GPULive tok/s in chatSpeed benchmarks9 inference engines

HuggingFace Ollama Library GGUF Downloads Build Hardware