RTX 4090 Cloud Rental Pricing
Consumer Ada GPU with 24GB, the cheapest fast option for small-model inference and Stable Diffusion.
7 offers from 7 providers · updated November 16, 2026
| Provider | Variant | VRAM | On-demand /hr | Spot /hr | Per month | Billing |
|---|---|---|---|---|---|---|
| Vast.ai | RTX 4090 | 24GB | $0.27 | — | $197 | per second |
| TensorDock | RTX 4090 | 24GB | $0.35 | — | $255 | per second |
| Salad | RTX 4090 | 24GB | $0.44 | $0.16 | $321 | per second |
| Oblivus | RTX 4090 | 24GB | $0.64 | — | $467 | per hour |
| RunPod credits | RTX 4090 | 24GB | $0.89 | — | $650 | per ms |
| Prime Intellect | RTX 4090 | 24GB | $0.92 | — | $672 | per minute |
| Fluence | RTX 4090 | 24GB | $1.06 | — | $774 | per hour |
Per GPU-hour, on-demand unless noted. Monthly = 730 hours. Check each provider for region availability and multi-GPU node minimums.
Or run models at home
If you rent a RTX 4090 mainly for inference, a mini PC with lots of unified memory can hold the same open models on your desk. It's slower per token than a RTX 4090, but there's no hourly bill.Compare mini PCs for local LLMs.
GMKtec EVO-X2
- Chip
- AMD Ryzen AI Max+ 395
- Memory
- 64GB or 128GB unified (up to 96GB as VRAM)
- Bandwidth
- 256 GB/s
- Runs
- 128GB: gpt-oss-120b, Qwen3.5 122B-A10B, GLM-4.5 Air. 64GB: 30B-class models
128GB / 1TB is $3,499.99. The cheapest 128GB Ryzen AI Max box we found.
Minisforum MS-S1 Max
- Chip
- AMD Ryzen AI Max+ 395
- Memory
- 128GB unified (up to 96GB as VRAM)
- Bandwidth
- 256 GB/s
- Runs
- gpt-oss-120b, Qwen3.5 122B-A10B, Llama 4 Scout, Mistral Small 4
Sale price (list $4,949) for 128GB / 2TB. Built as a small workstation, with more ports than most.
NVIDIA DGX Spark
- Chip
- NVIDIA GB10 Grace Blackwell
- Memory
- 128GB unified
- Bandwidth
- 273 GB/s
- Runs
- gpt-oss-120b, Qwen3.5 122B-A10B, Llama 4 Scout; CUDA, so most training tools work
128GB / 4TB Founders Edition. Partner GB10 boxes from ASUS, MSI and Dell cost $5,599 to $8,500.
Mac Studio (M5 Max)
- Chip
- Apple M5 Max
- Memory
- 48GB or 128GB unified
- Bandwidth
- 614 GB/s (40-core GPU)
- Runs
- 128GB: gpt-oss-120b, Qwen3.5 122B-A10B, at more than twice the bandwidth of the AMD boxes
128GB is $5,099. The M5 Ultra (96GB to 512GB, 1.2 TB/s) starts at $5,499.
Starting prices from each maker's own store on November 22, 2026; they change often, so check the store. "Runs" is a rough guide for 4-bit quantized models.
RTX 4090 guides
The Best Mini PCs for Running LLMs Locally in 2026
GMKtec EVO-X2, Minisforum MS-S1 Max, DGX Spark and Mac Studio for local LLMs: memory, bandwidth, which models fit, and store prices from $1,600.
Nov 2026Which Open Models Can You Run Locally? A Memory Guide for Qwen, DeepSeek, GLM and gpt-oss
How much RAM Qwen, GLM, DeepSeek, Llama and gpt-oss need at 4-bit, which mini PCs and Macs fit them, speed limits, and when local beats API prices.
RTX 4090 rental FAQ
How much does it cost to rent an RTX 4090?
On-demand RTX 4090 rental ranges from $0.27 to $1.06 per GPU-hour across 7 providers (median $0.64), as of November 16, 2026. Spot/interruptible capacity starts at $0.16/hr.
What is the cheapest RTX 4090 cloud provider?
Vast.ai has the lowest on-demand RTX 4090 price we track at $0.27 per GPU-hour (RTX 4090).
How much does an 8x RTX 4090 server cost per month?
At the cheapest on-demand rate ($0.27/GPU-hour), 8x RTX 4090 running 24/7 costs about $1,577 per month (730 hours). At the median rate it is about $3,738.