gaunernst

gemma-3-27b-it-int4-awq

image text to textgemma769K downloadsHugging Face

Cheapest GPU in stock that fits

A single GPU with at least 61 GB for inference.

NosanaIn stock · 21 freeStart right away

Nosana network, hosts worldwide

Details

Stock and price checked 11 min ago.

Self-service: you start it yourself, with no order or sales contact.

Confidential computing: data stays encrypted while the GPU works on it, so the host can't read it.

nvidia-pro-6000-cc

Full price

€2.09

per hour, whole instance, 1 GPU

Rule of thumb from the model's size. Real needs depend on batch size, context length and framework.

Inference (full precision)

61 GB

Inference (8-bit)

31 GB

Inference (4-bit)

15 GB

LoRA fine-tuning

66 GB

QLoRA (8-bit base)

33 GB

Full fine-tuning

245 GB