Skip to content

What can the RTX 3070 run?

What fits on the RTX 3070
Text models: 234Visual models: 6Strongest first
Memory
8 GB
Bandwidth
448 GB/s
Maker
NVIDIA

Open the calculator on this machineEvery model for 8 GB of VRAM

Text models that run on it

Text models that run on the RTX 3070, best first
ModelBest fitContextSpeedStrongest at
Qwen3.5 9B4-bit32K56 tok/sChat, Coding, Agents, Docs
Gemma 4 12B3-bit96K50 tok/sChat, Coding, Docs
Qwen3.5 4B8-bit60K59 tok/sAgents, Docs
Gemma 4 E4B6-bit28K67 tok/sChat, Coding, Docs
Ministral 3 8B3-bit20K71 tok/sDocs
Gemma 4 E2B8-bit128K96 tok/sChat
Nemotron 3 Nano 4B8-bit176K74 tok/sDocs
Llama 3.1 8B4-bit16K65 tok/sChat, Coding, Agents, Docs
Qwen3 0.6B8-bit40K295 tok/sNot rated
Qwen3 VL 8B Instruct3-bit16K73 tok/sText and images
Qwen3 8B3-bit20K78 tok/sNot rated
Qwen2.5 7B Instruct6-bit16K52 tok/sNot rated
Qwen2.5 VL 7B Instruct5-bit20K54 tok/sText and images
Unlimited OCR8-bit32K218 tok/sReading documents
Qwen2.5 1.5B Instruct8-bit32K165 tok/sNot rated
Qwen3.5 2B8-bit256K118 tok/sText and images
Qwen3 VL 4B Instruct8-bit16K62 tok/sText and images
Qwen3 1.7B8-bit40K129 tok/sNot rated
Qwen3.5 0.8B8-bit256K267 tok/sText and images
Qwen3 4B8-bit16K68 tok/sNot rated
Qwen3 4B Instruct 25078-bit16K68 tok/sNot rated
Qwen2.5 3B Instruct8-bit32K87 tok/sNot rated
DeepSeek OCR8-bit8K218 tok/sReading documents
Qwen2.5 VL 3B Instruct8-bit80K72 tok/sText and images
Qwen2.5 Coder 7B Instruct6-bit16K52 tok/sCoding
Mistral 7B Instruct V0.25-bit16K59 tok/sNot rated
Chandra OCR 28-bit40K52 tok/sReading documents
Llama 3.2 1B Instruct8-bit128K214 tok/sNot rated
Qwen3 VL 2B Instruct8-bit40K124 tok/sText and images
Qwen2 VL 2B Instruct8-bit32K120 tok/sText and images
MiniCPM5 2B8-bit104K113 tok/sNot rated
Gemma 3 4B IT8-bit112K75 tok/sText and images
Llama 3.2 3B Instruct8-bit32K85 tok/sNot rated
LightOnOCR 2 1B8-bit16K235 tok/sText and images
Qwen3 1.7B Base8-bit32K150 tok/sNot rated
HunyuanOCR8-bit60K222 tok/sText and images
Qwen3 0.6B Base8-bit32K350 tok/sNot rated
Rax 4.58-bit256K118 tok/sText and images
Qwen2 VL 7B Instruct5-bit20K54 tok/sText and images
Ornith 1.5 9B4-bit32K56 tok/sNot rated
Unlimited OCR 28-bit60K217 tok/sReading documents
Granite 4.1 3B8-bit44K79 tok/sNot rated
DeepSeek R1 Distill Qwen 1.5B8-bit128K165 tok/sNot rated
Mistral 7B V0.16-bit32K51 tok/sNot rated
Qwen2.5 Coder 3B Instruct8-bit32K87 tok/sCoding
Meta Llama 3 8B Instruct Abliterated V35-bit8K56 tok/sNot rated
EXAONE 3.5 7.8B Instruct4-bit16K66 tok/sNot rated
Saiga Llama3 8B5-bit8K56 tok/sNot rated
MiniCPM V 4.7 1B8-bit256K193 tok/sText and images
Qwen2.5 Coder 1.5B Instruct8-bit32K165 tok/sCoding
DeepSeek R1 0528 Qwen3 8B3-bit20K78 tok/sNot rated
Qwen3 4B Thinking 25078-bit16K68 tok/sNot rated
Phi 4 Mini Instruct8-bit20K72 tok/sNot rated
Dots.ocr8-bit128K97 tok/sReading documents
DeepSeek R1 Distill Qwen 7B6-bit16K52 tok/sNot rated
Phi 3 Mini 4K Instruct8-bit4K74 tok/sNot rated
DeepSeek OCR 28-bit8K210 tok/sReading documents
SmolLM3 3B8-bit52K87 tok/sNot rated
Qwen2 7B Instruct6-bit16K52 tok/sNot rated
Apertus 8B Instruct 25094-bit16K65 tok/sNot rated

Best fit is the highest quality, up to 8-bit, that leaves the use case’s context. Ranked by our use-case scores, then by Hugging Face downloads. Speeds are rough, plus or minus 25%. The top 60 of 234 are shown.

Image and video models

Image and video models that run on the RTX 3070
ModelBest fitPer outputStrongest at
Stable Diffusion XL 1.016-bit16 sImages
Z-Image Turbo8-bit23 sImages
FLUX.1 [dev]4-bit52 sImages
FLUX.3 Klein8-bit33 sImages
Stable Diffusion 3.5 Large5-bit3.7 minImages
Wan 2.2 TI2V-5B8-bit34 minVideo

Only just too big

Questions

What is the best LLM for the RTX 3070?

Of the open models we track, the strongest that run well on the RTX 3070 are Qwen3.5 9B, Gemma 4 12B, Qwen3.5 4B. 234 text models fit it in all.

Calculated from published specs, not benchmarked. All machines