Can I run it?

Pick your graphics card and see which AI models run on it, and how fast.

Updated 21 Oct 2026

Pick your graphics card

Using a Mac, a laptop or another card? Find hardware checks every machine on our list, Macs included.

Your graphics card

18 of 31 AI models run well on an RTX 3090 (24 GB of graphics memory). 13 don't fit.

An RTX 3090 costs about $1,370€1,224, used (median of sold listings).

Check RTX 3090 prices

Runs well18

The whole model fits in the graphics memory. Bigger models come first: they are usually smarter, and a little slower.

Qwen3.6-35B-A3BSpeed not estimated~23 GB
Qwen3 32BFast~25 words/s~23 GB
Qwen2.5 32BFast~25 words/s~23 GB
Gemma 4 31BFast~27 words/s~21 GB
Qwen3 Coder 30B A3BSpeed not estimated~20 GB
Gemma 2 27BFast~29 words/s~19 GB
Gemma 3 27BFast~31 words/s~18 GB
Qwen3.8-27BFast~32 words/s~18 GB
Muse Glimmer 30BFast~32 words/s~18 GB
OrcaCyber Nano 1.0Fast~33 words/s~17 GB
Mistral Small 3.2 24BFast~35 words/s~16 GB
gpt-oss-20bSpeed not estimated~13 GB
Phi-4 14BFast~51 words/s~11 GB
Qwen2.5 14BFast~52 words/s~11 GB
Gemma 2 9BFast~70 words/s~8.5 GB
Llama 3.1 8BFast~91 words/s~6.7 GB
Mistral 7B v0.3Fast~98 words/s~6.2 GB
Llama 3.2 3BFast~182 words/s~3.7 GB

Doesn't fit13

Not enough graphics memory on the card alone.

NVIDIA Nemotron 3.5 Bolt 74B-A9Bneeds ~45 GB
DeepSeek-R1-Distill 70Bneeds ~46 GB
Llama 3.3 70Bneeds ~46 GB
Kolibri-1needs ~51 GB
Qwen2.5 72Bneeds ~51 GB
gpt-oss-120bneeds ~64 GB
Llama 4 Scoutneeds ~68 GB
NVIDIA Nemotron 3.5 Super VLneeds ~73 GB
Ornith-1.5-122Bneeds ~77 GB
Qwen3.8-Flash-Nextneeds ~112 GB
Reflection Beam 1.0needs ~308 GB
DeepSeek V4.1 Flash 1017needs ~446 GB
Meta Muse Spark Openneeds ~472 GB

Every model on every card

The same answers in one table. Click a mark for the full answer, or a model for every card at once.

Runs
Runs slowly
Doesn't fit

Swipe the table sideways to see all 9 cards.

Which AI models run on which graphics card, at the standard quality (our estimate)
AI model	RTX 3060 12GB12 GB	RTX 4060 Ti 16GB16 GB	RTX 309024 GB	RTX 409024 GB	RTX 508016 GB	RTX 509032 GB	RX 7900 XTX24 GB	Dual RTX 309048 GB	Dual RTX 409048 GB
Llama 3.2 3Bneeds ~3.7 GB	✓	✓	✓	✓	✓	✓	✓	✓	✓
Llama 3.1 8Bneeds ~6.7 GB	✓	✓	✓	✓	✓	✓	✓	✓	✓
Mistral 7B v0.3needs ~6.2 GB	✓	✓	✓	✓	✓	✓	✓	✓	✓
Gemma 2 9Bneeds ~8.5 GB	✓	✓	✓	✓	✓	✓	✓	✓	✓
Qwen2.5 14Bneeds ~11 GB	✓	✓	✓	✓	✓	✓	✓	✓	✓
Phi-4 14Bneeds ~11 GB	✓	✓	✓	✓	✓	✓	✓	✓	✓
Gemma 2 27Bneeds ~19 GB	✗	✗	✓	✓	✗	✓	✓	✓	✓
Qwen2.5 32Bneeds ~23 GB	✗	✗	✓	✓	✗	✓	✓	✓	✓
Llama 3.3 70Bneeds ~46 GB	✗	✗	✗	✗	✗	✗	✗	✓	✓
Qwen2.5 72Bneeds ~51 GB	✗	✗	✗	✗	✗	✗	✗	~	~
DeepSeek-R1-Distill 70Bneeds ~46 GB	✗	✗	✗	✗	✗	✗	✗	✓	✓
NVIDIA Nemotron 3.5 Bolt 74B-A9Bneeds ~45 GB	✗	✗	✗	✗	✗	✗	✗	✓	✓
Reflection Beam 1.0needs ~308 GB	✗	✗	✗	✗	✗	✗	✗	✗	✗
NVIDIA Nemotron 3.5 Super VLneeds ~73 GB	✗	✗	✗	✗	✗	✗	✗	~	~
DeepSeek V4.1 Flash 1017needs ~446 GB	✗	✗	✗	✗	✗	✗	✗	✗	✗
Meta Muse Spark Openneeds ~472 GB	✗	✗	✗	✗	✗	✗	✗	✗	✗
Kolibri-1needs ~51 GB	✗	✗	✗	✗	✗	✗	✗	~	~
OrcaCyber Nano 1.0needs ~17 GB	✗	✗	✓	✓	✗	✓	✓	✓	✓
Qwen3.8-27Bneeds ~18 GB	✗	✗	✓	✓	✗	✓	✓	✓	✓
Qwen3.8-Flash-Nextneeds ~112 GB	✗	✗	✗	✗	✗	✗	✗	~	~
Qwen3.6-35B-A3Bneeds ~23 GB	✗	✗	✓	✓	✗	✓	✓	✓	✓
Muse Glimmer 30Bneeds ~18 GB	✗	✗	✓	✓	✗	✓	✓	✓	✓
Ornith-1.5-122Bneeds ~77 GB	✗	✗	✗	✗	✗	✗	✗	~	~
gpt-oss-20bneeds ~13 GB	✗	✓	✓	✓	✓	✓	✓	✓	✓
gpt-oss-120bneeds ~64 GB	✗	✗	✗	✗	✗	✗	✗	~	~
Gemma 3 27Bneeds ~18 GB	✗	✗	✓	✓	✗	✓	✓	✓	✓
Gemma 4 31Bneeds ~21 GB	✗	✗	✓	✓	✗	✓	✓	✓	✓
Qwen3 32Bneeds ~23 GB	✗	✗	✓	✓	✗	✓	✓	✓	✓
Qwen3 Coder 30B A3Bneeds ~20 GB	✗	✗	✓	✓	✗	✓	✓	✓	✓
Mistral Small 3.2 24Bneeds ~16 GB	✗	✗	✓	✓	✗	✓	✓	✓	✓
Llama 4 Scoutneeds ~68 GB	✗	✗	✗	✗	✗	✗	✗	~	~

Llama 3.3 70B on an RTX 4090 has its own written guide: Can I run Llama 3.3 70B on an RTX 4090?

Speeds and memory are our estimate from the hardware specs, not a measured test. Every answer is for the standard quality (Q4) and a conversation of about 12 pages of text (8,192 tokens). Prices excl. shipping. Euro prices incl. VAT. Dollar prices are converted from our euro prices without EU VAT, before US sales tax. How we calculate