self-hosted/ai
§01·spec · /gpus

RTX 3090 Ti

nvidia30 series24GB VRAM

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

§02·models that run on this GPU
30 total
LLM · 12
ModelBest speedMin VRAMWorksEvidence
Qwen3 30B-A3B166.9tokens/s24GB2benchesrecipecheck ↗
gpt-oss 20B160.3tokens/s24GB2benchesrecipecheck ↗
Qwen3-8B123.7tokens/s24GB2benchesrecipecheck ↗
Qwen3 14B76.2tokens/s24GB2benchesrecipecheck ↗
Qwen3 32B38tokens/s24GB2benchesrecipecheck ↗
DeepSeek R1 Distill 14B12GBrecipecheck ↗
Devstral Small 2 (24B)24GBrecipecheck ↗
Laguna XS 2.124GBrecipecheck ↗
Llama 3.1 8B10GBrecipecheck ↗
Mistral Small 3.2 24B24GBrecipecheck ↗
North Mini Code 1.024GBrecipecheck ↗
Ornith 1.0 35B24GBrecipecheck ↗
Multimodal · 2
ModelBest speedMin VRAMWorksEvidence
Gemma 4 E4B-IT20GBrecipecheck ↗
Gemma 4 26B MoE18GBrecipecheck ↗
Video · 6
3D · 1
ModelBest speedMin VRAMWorksEvidence
Waypoint 1.512GBrecipecheck ↗
§03·tested recipes
showing 6 of 30