GeForce RTX 4090
from $0.39/hr.
The cheapest way onto a GPU. 24 GB is enough for an 8B model in FP16, SDXL, or a weekend of experiments, at $0.39 an hour.
Billed from your wallet · stop any time
What the RTX 4090
is built for.
Tensor-core precisions: FP8, BF16, FP16, TF32.
Best at
- Serving 7B–8B models
- Stable Diffusion and ComfyUI
- Learning CUDA and prototyping
- Small fine-tunes with QLoRA
Launch from a terminal
Same card, same price as the console.
Will my model fit
on 24 GB?
Rules of thumb for weights plus a modest KV cache or optimizer state. Longer contexts and bigger batches need more.
| Workload | Needs | On a RTX 4090 |
|---|---|---|
| Llama 3.1 8B · FP16 inference | ~18 GB | Fits |
| Qwen 32B · FP8 inference | ~36 GB | Too big |
| Llama 70B · 4-bit inference | ~42 GB | Too big |
| Llama 70B · FP8 inference | ~78 GB | Too big |
| Llama 8B · LoRA fine-tune (BF16) | ~24 GB | Tight |
| 32B · QLoRA fine-tune | ~40 GB | Too big |
| Flux.1 [dev] · BF16 | ~33 GB | Too big |
| SDXL · FP16 | ~12 GB | Fits |
Guides for
the RTX 4090.
Benchmarks and sizing guides from our blog.
Rent RTX 4090 in the Cloud – Deploy in Minutes on packet.ai
The fastest consumer GPU ever made, single-tenant and dedicated for $0.39/hr. 24 GB GDDR6X for 7–13B inference, SDXL, embeddings and rapid experimentation.
RTX 4090 for AI in the Cloud: Pricing, Performance and How It Compares to A100
24 GB GDDR6X, Ada Lovelace, from $0.39/hr on packet.ai. The RTX 4090 is the cheapest capable GPU for AI. Here is exactly what it runs, what it cannot, and when to upgrade.
Not the right fit?
Compare the fleet.
Every card on sale is billed hourly, with no contract.
Launch a RTX 4090
from $0.39/hr.
Sign up, top up, launch. You'll be on a GPU in minutes.
