Skip to content

Models for Local GPU

Models for Local GPU

VRAM Size Category Models
4 GB Text llama3.2 3b
Vision moondream, llava-phi3
Speech Whisper - tiny
6 GB Text llama3.1 8b, mistral7b, ministral8b, qwen2.5
Vision llava, minicpm-v
Speech Whisper - medium
12 GB Text mistral-nemo
Vision llava
16 GB Text
Vision llama3.2 vision