§01·spec · /gpus

RTX 4070

Name: RTX 4070 local AI benchmarks
Creator: Smeltcore
License: https://creativecommons.org/licenses/by-sa/4.0/

nvidia40 series12GB VRAM

37 open-weights AI models run on the RTX 4070 — see which fit, how fast they go, and the VRAM each needs.

§02·models that run on this GPU

37 total

LLM · 9

Model	Min VRAM	Works	Evidence
Gemma 4 12B	12GB	✓	recipe	check ↗
gpt-oss 20B	12GB	✓	recipe	check ↗
Llama 3.1 8B	10GB	✓	recipe	check ↗
Mistral Nemo 12B	12GB	✓	recipe	check ↗
Ornith 1.0 9B	12GB	✓	recipe	check ↗
Phi-4	12GB	✓	recipe	check ↗
Qwen3 14B	12GB	✓	recipe	check ↗
Qwen3 30B-A3B	12GB	✓	recipe	check ↗
Qwen3-8B	12GB	✓	recipe	check ↗

Multimodal · 3

Model	Best speed	Min VRAM	Works	Evidence
Qwen3.6 35B-A3B	80.8tokens/s	12GB	✓	1benchrecipe	check ↗
Gemma 4 E4B-IT		6GB	✓	recipe	check ↗
MiniMind-O		4GB	✓	recipe	check ↗

Image · 9

Model	Min VRAM	Works	Evidence
Anima	7GB	✓	recipe	check ↗
Chroma V48	11GB	✓	recipe	check ↗
ERNIE-Image-Turbo	12GB	✓	recipe	check ↗
Flux.2-Klein-4B	8GB	✓	recipe	check ↗
HiDream-O1-Image	10GB	✓	recipe	check ↗
Juggernaut Z	12GB	✓	recipe	check ↗
Krea 2	12GB	✓	recipe	check ↗
Qwen-Image	12GB	✓	recipe	check ↗
SenseNova U1	12GB	✓	recipe	check ↗

Video · 3

Model	Min VRAM	Works	Evidence
CogVideoX 1.5	10GB	✓	recipe	check ↗
LightX2V	8GB	✓	recipe	check ↗
WAN 2.2	8GB	✓	recipe	check ↗

3D · 1

Model	Best speed	Min VRAM	Works	Evidence
Hunyuan3D		10GB	✓	recipe	check ↗

TTS · 10

Model	Min VRAM	Works	Evidence
ACE-Step 1.5 XL	8GB	✓	recipe	check ↗
Foundation-1	8GB	✓	recipe	check ↗
Kokoro TTS	2GB	✓	recipe	check ↗
MOSS-Audio	12GB	✓	recipe	check ↗
OmniVoice	4GB	✓	recipe	check ↗
OpenAudio S1 Mini	5GB	✓	recipe	check ↗
Qwen3-TTS	8GB	✓	recipe	check ↗
VoxCPM	5GB	✓	recipe	check ↗
VoxCPM2	8GB	✓	recipe	check ↗
Voxtral Mini 3B	10GB	✓	recipe	check ↗

specialized · 2

Model	Best speed	Min VRAM	Works	Evidence
KiMoDo		3GB	✓	recipe	check ↗
SAM 3		4GB	✓	recipe	check ↗

§03·tested recipes

showing 6 of 37