GPU EconomyCloud GPU prices, read every hour
Latest read 115 providers tracked27,573 prices read in 24 h
H100 index$3.2736 providersWidestV100 32GB20.0×Mover, 7dL40+7.2%

Marketplace · 59 GPU models tracked

Vast.ai

$0.05–$10.63per GPU-hour, the cheapest row to the dearest

A marketplace where hosts set their own prices. Each figure is a floor - the median of the five cheapest verified single-GPU offers - withheld when too few offers are listed. Machines with several GPUs are a separate row.

As of , Vast.ai lists 59 GPU models we track, from $0.05 per GPU-hour (NVIDIA GeForce GTX 1080, community, floor) to $10.63 (NVIDIA B200, community, floor).

Every rate ↓ · Compare with 93 providers ↓

  • Cheapest row$0.05NVIDIA GeForce GTX 1080 · community, floor
  • B200$10.63marketplace floor · no on-demand rate
  • H200$4.94marketplace floor · no on-demand rate
  • A100 80GB$1.00marketplace floor · no on-demand rate
Websitevast.ai
Pricing pagecloud.vast.ai
Regions and tiersMulti-GPU machines — as labelled in the rates below
Models tracked59

Price by GPU

As of , per GPU-hour: the rate Vast.ai lists for each card, with the market it is sold in. Vast.ai lists no on-demand rate for any card we track, so no card here is placed against the providers’ on-demand rate cards; each card’s page shows its rates beside theirs.

GPUVast.ai’s listed rate, per GPU-hour
NVIDIA B200$10.63 marketplace floor
captured
Deploy ↗
NVIDIA H200$4.94 marketplace floor
captured
Deploy ↗
NVIDIA H200 NVL$4.00 marketplace floor
captured
Deploy ↗
NVIDIA H100 PCIe$2.85 marketplace floor
captured
Deploy ↗
NVIDIA RTX PRO 6000 Blackwell Max-Q$1.55 marketplace floor
captured
Deploy ↗
NVIDIA RTX PRO 6000 Blackwell$1.34 marketplace floor
captured
Deploy ↗
NVIDIA A100 80GB$1.00 marketplace floor
captured
Deploy ↗
NVIDIA L40S$0.73 marketplace floor
captured
Deploy ↗
NVIDIA L4$0.67 marketplace floor
captured
Deploy ↗
NVIDIA RTX PRO 5000 Blackwell$0.62 marketplace floor
captured
Deploy ↗
NVIDIA RTX 6000 Ada$0.59 marketplace floor
captured
Deploy ↗
NVIDIA L40$0.51 marketplace floor
captured
Deploy ↗
NVIDIA RTX 5000 Ada Generation$0.48 marketplace floor
captured
Deploy ↗
NVIDIA RTX 5090$0.47 marketplace floor
captured
Deploy ↗
NVIDIA RTX A6000$0.46 marketplace floor
captured
Deploy ↗
NVIDIA A100 40GB$0.44 marketplace floor
captured
Deploy ↗
NVIDIA RTX 4090$0.37 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 3090 Ti$0.36 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 4080 SUPER$0.27 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 4070 Ti SUPER$0.24 marketplace floor
captured
Deploy ↗
NVIDIA RTX A5000$0.23 marketplace floor
captured
Deploy ↗
NVIDIA RTX 4000 Ada$0.22 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 5080$0.22 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 4080$0.21 marketplace floor
captured
Deploy ↗
NVIDIA TITAN RTX$0.21 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 5070$0.18 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 4070 SUPER$0.16 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 5070 Ti$0.16 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 3090$0.16 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 4070$0.13 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 4060 Ti$0.12 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 2080 Ti$0.12 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 3080$0.12 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 3080 Ti$0.12 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 5060 Ti$0.12 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 5060$0.11 marketplace floor
captured
Deploy ↗
NVIDIA RTX A4000$0.09 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 3070$0.08 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 4060$0.08 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 3060 Ti$0.07 marketplace floor
captured
Deploy ↗
NVIDIA Quadro P4000$0.07 marketplace floor
captured
Deploy ↗
NVIDIA GeForce GTX 1080 Ti$0.06 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 2060 12GB$0.06 marketplace floor
captured
Deploy ↗
NVIDIA GeForce RTX 3060$0.06 marketplace floor
captured
Deploy ↗
NVIDIA GeForce GTX 1070 Ti$0.06 marketplace floor
captured
Deploy ↗
NVIDIA GeForce GTX 1660 SUPER$0.06 marketplace floor
captured
Deploy ↗
NVIDIA TITAN Xp$0.05 marketplace floor
captured
Deploy ↗
NVIDIA GeForce GTX 1080$0.05 marketplace floor
captured
Deploy ↗

The same, in words

All listed prices

One table per price type, never mixed. Every rate normalized to one GPU for one hour; a price opens the page it was read from, and 7d marks a rate that moved in the past seven days. The day each figure was first listed is in this page’s data, below.

On-demand

Listed pay-as-you-go rates at Vast.ai, cheapest first.
GPU · what is sold$ / GPU-hourReadSource
NVIDIA GeForce GTX 1080floorAPIcommunity · $0.10/node ÷ 2 · Multi-GPU machines · 7d −5.0%$0.05captured Deploy ↗
NVIDIA TITAN XpfloorAPIcommunity · $0.11/node ÷ 2 · Multi-GPU machines · 7d −18.7%$0.05captured Deploy ↗
NVIDIA Quadro P4000floorAPIcommunity · $0.11/node ÷ 2 · Multi-GPU machines$0.06thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce GTX 1080floorAPIcommunity · 7d +5.1%$0.06captured Deploy ↗
NVIDIA GeForce GTX 1660 SUPERfloorAPIcommunity · $0.11/node ÷ 2 · Multi-GPU machines$0.06captured Deploy ↗
NVIDIA GeForce GTX 1660 SUPERfloorAPIcommunity$0.06captured Deploy ↗
NVIDIA GeForce GTX 1070 TifloorAPIcommunity · $0.24/node ÷ 4 · Multi-GPU machines · 7d +12.4%$0.06captured Deploy ↗
NVIDIA GeForce RTX 3060floorAPIcommunity · $0.12/node ÷ 2 · Multi-GPU machines · 7d +0.2%$0.06captured Deploy ↗
NVIDIA GeForce RTX 2060 12GBfloorAPIcommunity · $0.12/node ÷ 2 · Multi-GPU machines · 7d −24.7%$0.06captured Deploy ↗
NVIDIA GeForce RTX 3060floorAPIcommunity · 7d −10.7%$0.06captured Deploy ↗
NVIDIA GeForce RTX 2060 12GBfloorAPIcommunity · 7d −24.3%$0.06captured Deploy ↗
NVIDIA TITAN XpfloorAPIcommunity · 7d −9.7%$0.06captured Deploy ↗
NVIDIA GeForce GTX 1080 TifloorAPIcommunity · 7d −5.9%$0.06captured Deploy ↗
NVIDIA GeForce GTX 1080 TifloorAPIcommunity · $0.14/node ÷ 2 · Multi-GPU machines · 7d +7.3%$0.07captured Deploy ↗
NVIDIA GeForce RTX 3060 TifloorAPIcommunity · 7d −40.4%$0.07captured Deploy ↗
NVIDIA Quadro P4000floorAPIcommunity$0.07captured Deploy ↗
NVIDIA GeForce RTX 4060floorAPIcommunity · $0.31/node ÷ 4 · Multi-GPU machines · 7d −3.8%$0.08captured Deploy ↗
NVIDIA Tesla P100 12GBfloorAPIcommunity · $0.32/node ÷ 4 · Multi-GPU machines$0.08thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce GTX 1660 TifloorAPIcommunity$0.08thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce RTX 3070floorAPIcommunity · $0.16/node ÷ 2 · Multi-GPU machines · 7d −49.2%$0.08captured Deploy ↗
NVIDIA GeForce RTX 4060floorAPIcommunity · 7d +3.0%$0.08captured Deploy ↗
NVIDIA RTX A4000floorAPIcommunity · $0.17/node ÷ 2 · Multi-GPU machines · 7d −27.8%$0.09captured Deploy ↗
NVIDIA RTX A4000floorAPIcommunity · 7d −20.2%$0.09captured Deploy ↗
NVIDIA GeForce RTX 3060 TifloorAPIcommunity · $0.19/node ÷ 2 · Multi-GPU machines · 7d −42.1%$0.09captured Deploy ↗
NVIDIA V100floorAPIcommunity · $0.19/node ÷ 2 · Multi-GPU machines$0.09thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce RTX 3070floorAPIcommunity · 7d −21.3%$0.10captured Deploy ↗
NVIDIA Quadro RTX 4000floorAPIcommunity$0.11thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce RTX 5060floorAPIcommunity · 7d −10.2%$0.11captured Deploy ↗
NVIDIA GeForce RTX 5060 TifloorAPIcommunity · $0.24/node ÷ 2 · Multi-GPU machines · 7d −5.7%$0.12captured Deploy ↗
NVIDIA GeForce RTX 5060 TifloorAPIcommunity · 7d −6.1%$0.12captured Deploy ↗
NVIDIA GeForce RTX 3080 TifloorAPIcommunity · $0.24/node ÷ 2 · Multi-GPU machines · 7d −39.9%$0.12captured Deploy ↗
NVIDIA GeForce RTX 3080 TifloorAPIcommunity · 7d −26.7%$0.12captured Deploy ↗
NVIDIA GeForce RTX 2080 TifloorAPIcommunity · $0.24/node ÷ 2 · Multi-GPU machines$0.12thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce RTX 3080floorAPIcommunity · 7d −24.4%$0.12captured Deploy ↗
NVIDIA GeForce RTX 2080 TifloorAPIcommunity$0.12captured Deploy ↗
NVIDIA V100floorAPIcommunity$0.12thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce RTX 4060 TifloorAPIcommunity · $0.25/node ÷ 2 · Multi-GPU machines · 7d −7.1%$0.12captured Deploy ↗
NVIDIA GeForce RTX 4070floorAPIcommunity · 7d −25.7%$0.13captured Deploy ↗
NVIDIA GeForce RTX 4060 TifloorAPIcommunity · 7d +11.2%$0.13captured Deploy ↗
NVIDIA GeForce RTX 3080floorAPIcommunity · $0.54/node ÷ 4 · Multi-GPU machines · 7d −33.0%$0.13captured Deploy ↗
NVIDIA RTX A2000floorAPIcommunity · $0.31/node ÷ 2 · Multi-GPU machines$0.15thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce RTX 3090floorAPIcommunity · $1.28/node ÷ 8 · Multi-GPU machines · 7d −8.6%$0.16captured Deploy ↗
NVIDIA GeForce RTX 5070 TifloorAPIcommunity · $0.32/node ÷ 2 · Multi-GPU machines · 7d −47.7%$0.16captured Deploy ↗
NVIDIA GeForce RTX 5070 TifloorAPIcommunity · 7d −19.8%$0.16captured Deploy ↗
NVIDIA GeForce RTX 3070 TifloorAPIcommunity$0.16thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce RTX 4070 SUPERfloorAPIcommunity · 7d −10.5%$0.16captured Deploy ↗
NVIDIA GeForce RTX 5070floorAPIcommunity · 7d −13.5%$0.18captured Deploy ↗
NVIDIA GeForce RTX 4070 SUPERfloorAPIcommunity · $0.38/node ÷ 2 · Multi-GPU machines · 7d +15.6%$0.19captured Deploy ↗
NVIDIA GeForce RTX 3090floorAPIcommunity · 7d +6.7%$0.19captured Deploy ↗
NVIDIA GeForce RTX 4070floorAPIcommunity · $1.60/node ÷ 8 · Multi-GPU machines · 7d −0.4%$0.20captured Deploy ↗
NVIDIA TITAN VfloorAPIcommunity$0.20thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA RTX A2000floorAPIcommunity$0.20thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA TITAN RTXfloorAPIcommunity · 7d −11.8%$0.21captured Deploy ↗
NVIDIA GeForce RTX 4080floorAPIcommunity · $0.43/node ÷ 2 · Multi-GPU machines · 7d −40.2%$0.21captured Deploy ↗
NVIDIA GeForce RTX 5080floorAPIcommunity · $0.86/node ÷ 4 · Multi-GPU machines · 7d −35.4%$0.22captured Deploy ↗
NVIDIA RTX 4000 AdafloorAPIcommunity · 7d −8.0%$0.22captured Deploy ↗
NVIDIA GeForce RTX 5080floorAPIcommunity · 7d −40.6%$0.22captured Deploy ↗
NVIDIA TITAN RTXfloorAPIcommunity · $0.45/node ÷ 2 · Multi-GPU machines$0.22thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce RTX 3090 TifloorAPIcommunity · $0.46/node ÷ 2 · Multi-GPU machines$0.23thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA RTX A5000floorAPIcommunity · $0.46/node ÷ 2 · Multi-GPU machines · 7d +0.3%$0.23captured Deploy ↗
NVIDIA RTX A5000floorAPIcommunity · 7d +1.4%$0.23captured Deploy ↗
NVIDIA RTX 4000 AdafloorAPIcommunity · $0.96/node ÷ 4 · Multi-GPU machines$0.24thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce RTX 4080floorAPIcommunity · 7d −9.9%$0.24captured Deploy ↗
NVIDIA GeForce RTX 4070 Ti SUPERfloorAPIcommunity · 7d +2.8%$0.24captured Deploy ↗
NVIDIA GeForce RTX 5070floorAPIcommunity · $0.50/node ÷ 2 · Multi-GPU machines · 7d +16.4%$0.25captured Deploy ↗
NVIDIA GeForce RTX 4080 SUPERfloorAPIcommunity · 7d +0.8%$0.27captured Deploy ↗
NVIDIA GeForce RTX 4070 Ti SUPERfloorAPIcommunity · $1.12/node ÷ 4 · Multi-GPU machines · 7d +16.4%$0.28captured Deploy ↗
NVIDIA GeForce RTX 4080 SUPERfloorAPIcommunity · $1.33/node ÷ 4 · Multi-GPU machines · 7d −4.5%$0.33captured Deploy ↗
NVIDIA RTX 5000 Ada GenerationfloorAPIcommunity · $0.72/node ÷ 2 · Multi-GPU machines$0.36thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA GeForce RTX 3090 TifloorAPIcommunity · 7d +26.1%$0.36captured Deploy ↗
NVIDIA RTX 4090floorAPIcommunity · $0.75/node ÷ 2 · Multi-GPU machines · 7d −4.0%$0.37captured Deploy ↗
NVIDIA RTX 4090floorAPIcommunity · 7d −10.3%$0.39captured Deploy ↗
NVIDIA A100 40GBfloorAPIcommunity · $0.88/node ÷ 2 · Multi-GPU machines · 7d −30.2%$0.44captured Deploy ↗
NVIDIA RTX A6000floorAPIcommunity · $0.91/node ÷ 2 · Multi-GPU machines · 7d −5.5%$0.46captured Deploy ↗
NVIDIA RTX A6000floorAPIcommunity · 7d +6.2%$0.46captured Deploy ↗
NVIDIA RTX 5090floorAPIcommunity · $0.93/node ÷ 2 · Multi-GPU machines · 7d −21.4%$0.47captured Deploy ↗
NVIDIA RTX 4500 Ada GenerationfloorAPIcommunity · $0.94/node ÷ 2 · Multi-GPU machines$0.47thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA RTX 4500 Ada GenerationfloorAPIcommunity$0.47thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA RTX 5000 Ada GenerationfloorAPIcommunity · 7d +8.5%$0.48captured Deploy ↗
NVIDIA RTX 5090floorAPIcommunity · 7d −8.5%$0.50captured Deploy ↗
NVIDIA L40floorAPIcommunity · 7d +11.2%$0.51captured Deploy ↗
NVIDIA A100 40GBfloorAPIcommunity · 7d −18.5%$0.58captured Deploy ↗
NVIDIA RTX 6000 AdafloorAPIcommunity · 7d −5.8%$0.59captured Deploy ↗
NVIDIA RTX 6000 AdafloorAPIcommunity · $1.22/node ÷ 2 · Multi-GPU machines · 7d −8.3%$0.61captured Deploy ↗
NVIDIA RTX PRO 5000 BlackwellfloorAPIcommunity · 7d −7.1%$0.62captured Deploy ↗
NVIDIA L4floorAPIcommunity · $1.34/node ÷ 2 · Multi-GPU machines$0.67captured Deploy ↗
NVIDIA L4floorAPIcommunity · 7d +35.0%$0.67captured Deploy ↗
NVIDIA L40SfloorAPIcommunity · $1.47/node ÷ 2 · Multi-GPU machines$0.73captured Deploy ↗
NVIDIA L40SfloorAPIcommunity$0.73captured Deploy ↗
NVIDIA L40floorAPIcommunity · $3.15/node ÷ 4 · Multi-GPU machines$0.79thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA RTX 5880 AdafloorAPIcommunity · $1.60/node ÷ 2 · Multi-GPU machines$0.80thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA RTX 5880 AdafloorAPIcommunity$0.80thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA RTX PRO 5000 BlackwellfloorAPIcommunity · $1.74/node ÷ 2 · Multi-GPU machines · 7d +18.3%$0.87captured Deploy ↗
NVIDIA A100 80GBfloorAPIcommunity · $4.00/node ÷ 4 · Multi-GPU machines · 7d −9.4%$1.00captured Deploy ↗
NVIDIA A100 80GBfloorAPIcommunity · 7d +2.4%$1.14captured Deploy ↗
NVIDIA RTX PRO 6000 BlackwellfloorAPIcommunity · 7d +0.1%$1.34captured Deploy ↗
NVIDIA RTX PRO 6000 BlackwellfloorAPIcommunity · $5.87/node ÷ 4 · Multi-GPU machines · 7d +3.8%$1.47captured Deploy ↗
NVIDIA RTX PRO 6000 Blackwell Max-QfloorAPIcommunity · 7d +30.3%$1.55captured Deploy ↗
NVIDIA RTX PRO 6000 Blackwell Max-QfloorAPIcommunity · $3.50/node ÷ 2 · Multi-GPU machines · 7d +5.0%$1.75captured Deploy ↗
NVIDIA H100 SXMfloorAPIcommunity$2.27thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA H100 PCIefloorAPIcommunity$2.85captured Deploy ↗
NVIDIA H100 NVLfloorAPIcommunity · $12.48/node ÷ 4 · Multi-GPU machines$3.12thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA H100 PCIefloorAPIcommunity · $12.91/node ÷ 4 · Multi-GPU machines$3.23thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA H100 NVLfloorAPIcommunity$3.30thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA H200 NVLfloorAPIcommunity · $8.00/node ÷ 2 · Multi-GPU machines$4.00captured Deploy ↗
NVIDIA H200 NVLfloorAPIcommunity · 7d +3.2%$4.00captured Deploy ↗
NVIDIA H100 SXMfloorAPIcommunity · $8.51/node ÷ 2 · Multi-GPU machines$4.26thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA H200floorAPIcommunity · $39.48/node ÷ 8 · Multi-GPU machines · 7d −3.8%$4.94captured Deploy ↗
NVIDIA H200floorAPIcommunity$5.28thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA B200floorAPIcommunity$7.50thin marketfewer offers than a floor needs · last floor Deploy ↗
NVIDIA B200floorAPIcommunity · $42.51/node ÷ 4 · Multi-GPU machines · 7d +41.7%$10.63captured Deploy ↗

Ask an AI agent about Vast.ai’s rates: Claude, Cursor, VS Code and ChatGPT can read them through the GPU Economy MCP server - free, no key, every answer citing the page and read time.

What else Vast.ai bills

The GPU-hour is not the whole bill. What Vast.ai's own pages say about the rest, read by hand and checked on ; each line links the page it was read from. "Not published" means we looked and the pages do not say. Every provider side by side.

WhatWhat Vast.ai statesSource
In the rateGPU rate includes a GPU-proportional share of host CPU and RAM; storage and bandwidth priced separatelysource
checked
Data outSet by each host per offer ($/TB), charged on every byte up and down, in any instance state; no platform-wide ratesource
checked
StorageSet by each host per offer, charged on allocated disk size; no platform-wide pricesource
checked
While stoppedGPU rental stops; storage keeps billing until the instance is deleted (bandwidth billed in any state)source
checked
Billing stepPer second while active; credits must be prepaid; no minimum charge publishedsource
checked

Referral link: Sign up with Vast.ai referral link The Deploy buttons for Vast.ai on this site go through this link. We may earn a commission if you sign up through it. It changes no price, no order and no inclusion on this site: link policy. Every price itself links to its source directly.

Every rate on this page as data: /providers/vast-ai/prices.json — the same fields as prices.json, narrowed to this provider. CC BY 4.0, no key, no rate limit. On your own page: embed this table (/embed/providers/vast-ai, an iframe, no script).

Compare Vast.ai with

Every provider that lists at least two of the same GPUs on-demand, most shared GPUs first; each opens the side-by-side page.

Data updated: