self-hosted/ai
§01·spec · /gpus

RTX 3090

nvidia30 series24GB VRAM

35 open-weights AI models run on the RTX 3090 — see which fit, how fast they go, and the VRAM each needs.

§02·models that run on this GPU
35 total
Multimodal · 5
ModelBest speedMin VRAMWorksEvidence
Gemma 4 26B MoE119.4tokens/s24GB2benchesrecipecheck ↗
Qwen3.5 35B111.2tokens/s24GB2benchesrecipecheck ↗
Gemma4 31B34.7tokens/s24GB2benchesrecipecheck ↗
Qwen3.5 27B33.5tokens/s24GB2benchesrecipecheck ↗
Gemma 4 E4B-IT20GBrecipecheck ↗
Image · 8
Video · 6
ModelBest speedMin VRAMWorksEvidence
Wan 2.2 14B7.17min24GB1benchrecipecheck ↗
CogVideoX 1.514GBrecipecheck ↗
HunyuanVideo 1.524GBrecipecheck ↗
LightX2V14GBrecipecheck ↗
Mochi 122GBrecipecheck ↗
WAN 2.224GBrecipecheck ↗
3D · 1
ModelBest speedMin VRAMWorksEvidence
Waypoint 1.512GBrecipecheck ↗
§03·tested recipes
showing 6 of 35