About NVIDIA RTX 5090
The NVIDIA RTX 5090 rental price today is $0.70 per GPU-hour, up 0.2% since yesterday, up 2.3% over the past week and down 0.1% over the past month. At that rate one RTX 5090 costs about $510 for a month of 730 hours. Over the past year it has ranged from $0.39 to $0.73. The index is the price of one benchmark contract: GeForce RTX 5090 32GB, on-demand, 8-GPU node, US hardware, CPU-neutral. It is standardized across 5 cloud providers and marketplaces, from Akash at $0.58 to RunPod at $1.04, and is updated every minute.
The GeForce RTX 5090 is the flagship of NVIDIA's consumer GeForce RTX 50 Series, announced on January 6, 2025 for sale from January 30 at $1,999. It is a Blackwell PCIe card with 32 GB of GDDR7 memory and no NVLink, and the successor of the RTX 4090. Many GPU clouds and marketplaces rent it: on November 10, 2026 GetDeploying tracked 22 RTX 5090 providers, 17 of them with an on-demand price.
Specifications
| Spec | RTX 5090 |
|---|
| Architecture | Blackwell |
| Cores | 21,760 CUDA, fifth-generation Tensor, fourth-generation RT |
| AI performance | 3,352 AI TOPS, as NVIDIA states it |
| Ray tracing | 318 TFLOPS |
| Memory | 32 GB GDDR7, 512-bit, 1,792 GB/s |
| Boost clock | 2.41 GHz |
| Interface | PCIe Gen 5; no NVLink |
| Power | 575 W total graphics power; Founders Edition 2 slots, 304 mm long |
| Video | 3 ninth-generation encoders and 2 sixth-generation decoders |
Source: NVIDIA GeForce RTX 5090, accessed November 11, 2026; memory bandwidth from NVIDIA's Blackwell whitepaper.
Where the RTX 5090 is rented
GetDeploying tracked 22 RTX 5090 providers on November 10, 2026, among them RunPod, Salad, Novita, Lium, Oblivus and Sesterce, and 17 of them had an on-demand price; none was AWS, Azure, Google Cloud or Oracle.
What this price measures
The index is the posted on-demand price of one benchmark contract, a GeForce RTX 5090 32GB in an 8-GPU node of US hardware, rented by the hour from GPU clouds and marketplaces. Prices for other node sizes are converted to that contract with the index's node-size factors, and prices from other regions are converted with the index's region factors. Reserved and spot prices are not counted.
On November 10, 2026, GetDeploying listed 8-GPU RTX 5090 machines on demand at $0.71 per GPU-hour at Sesterce, out of stock, and $0.72 at Runcrate, in stock. Its medians were $0.72 on demand, across 60 listings, $0.62 reserved and $0.65 spot.
What moves RTX 5090 rental prices
- Inference that fits on one card. One RTX 5090 can run models up to about 41 billion parameters at 4-bit precision, or 9 billion at 16-bit, with a 32K context, GetDeploying estimates. With no NVLink, cards in one machine exchange data over PCIe Gen 5.
- Export controls. For China, NVIDIA sells the RTX 5090 D v2, "designed to comply with U.S. export controls", with less memory and bandwidth than the RTX 5090. China reportedly added it to a list of goods banned at its customs on May 15, 2026, according to the Financial Times as Tom's Hardware reported on May 20, 2026.
- Capacity. Where a source shows capacity, this index counts a provider only on days when one of its offers had capacity; its posted prices then all count, so an offer going out of stock does not read as a price change (methodology).
Price history
This index's history before November 6, 2026, its first day, is GetDeploying's weekly RTX 5090 series, rescaled to join the daily index, counting only weeks with at least 5 providers; it starts with the week of November 6, 2025.
Futures and benchmarks
No futures contract or Bloomberg ticker covers the RTX 5090 yet. The GPU futures ICE announced in May 2026 may cover the H100, H200, B200 and RTX 5090 and other GPUs "as the market develops". CME's filed contracts and the contracts ICE says it will list, with no date set yet, cover the H100 and B200.
FAQ
How does it compare with the RTX 4090? The RTX 5090 has 32 GB of GDDR7 against the RTX 4090's 24 GB of GDDR6X, 1,792 GB/s of memory bandwidth against 1,008 GB/s, and 21,760 CUDA cores against 16,384, at 575 W against 450 W, according to NVIDIA's RTX 5090 and RTX 4090 pages and its Blackwell and Ada whitepapers.
Why is a consumer card rented in clouds? Its Blackwell Tensor Cores support FP4, according to NVIDIA's Blackwell whitepaper, and GPU marketplaces rent it by the hour (above). NVIDIA's driver license, in the version of February 25, 2025, says GeForce software "is not licensed for datacenter deployment"; none of AWS, Azure, Google Cloud or Oracle was among the card's providers on GetDeploying (above).
How often is this index updated? It is recalculated every minute, with one value per day in US Central time that is final when the day ends.
Sources
- NVIDIA: GeForce RTX 5090 product page and specifications
- NVIDIA: GeForce RTX 50 Series announcement, January 6, 2025
- NVIDIA: RTX Blackwell GPU architecture whitepaper
- NVIDIA: driver license agreement, version of February 25, 2025
- GetDeploying: NVIDIA RTX 5090 prices
- GetDeploying: GPU price dataset, weekly price history, CC BY 4.0
- Tom's Hardware: China banned Nvidia's 5090D V2, May 20, 2026
- Business Wire: ICE and Ornn to launch GPU compute futures, May 19, 2026
Text reviewed Nov 11, 2026; the figures in it are as of that date. The figures at the top of this page update every minute.
Cite as: Compute Screener, “NVIDIA RTX 5090 rental price index”, computescreener.com/gpu/rtx-5090. The figures are free to reuse under CC BY 4.0 with credit to Compute Screener.