Hinode: cloud GPU prices and features, compared
Hinode next to RunPod, Vast.ai and Google Colab: what an hour costs, what is still there after you stop, and how you get in. Their figures come from their own pages, with the date.
RunPod · Vast.ai · Google Colab ·
Hinode prices are the live list
Price per hour
Hinode costs more per hour. Here is how much.
One GPU, on demand, in US dollars. The Hinode column is the live price list; the others are public list prices read on 2026-10-05.
| Hinode | RunPod · Secure | RunPod · Community | Vast.ai | Google Colab | |
| 24 GB | L4 · from $1.75 | L4 · $0.49 | RTX 4090 · $0.34 | RTX 4090 · about $0.50 | No hourly rate |
| 48 GB | L40S · from $3.30 | L40S · $1.09 | L40S · $0.79 | L40S · about $0.80 | Not sold by the card |
| 96 GB | RTX PRO 6000 · from $8.05 | RTX PRO 6000 · $2.09 | RTX PRO 6000 · $1.69 | None listed that day | Not sold by the card |
| Billed | By the minute | By the second | By the second | By the second | Compute units on a plan |
| Who sets the price | One list, both regions | RunPod | RunPod | Each host, its own |
Swipe for Vast.ai and Google Colab →
Vast.ai is a marketplace: its figures are the median of verified, on-demand, one-GPU offers on 2026-10-05, and they move by the hour. A Hinode rate is the whole machine: the card, its vCPUs and RAM. Disk and traffic are below.
Disk and traffic
What you pay when the GPU is off
A stopped machine still holds your files, and every provider bills for that differently.
| Hinode | RunPod | Vast.ai | Google Colab | |
| Disk, running | $0.15 per GB-month, for what is written | $0.10 per GB-month | Set by the host, about $0.27 per GB-month | Included |
| Disk, stopped | $0.15 per GB-month, same | $0.20 per GB-month | Still billed until you destroy it | Nothing is kept |
| Shared storage | $0.10 per GB-month, on every machine | $0.07 per GB-month network volume | Not checked | Google Drive |
| Traffic out | $0.10 per GB | No fee | Set by the host | Not checked |
Swipe for Vast.ai and Google Colab →
RunPod bills a stopped Pod's volume at twice the running rate. Vast.ai: median of the same offers as above.
What you get
A rented card, or a computer you come back to
Where the hour is cheaper, more of the setup is yours to keep alive.
| Hinode | RunPod | Vast.ai | Google Colab | |
| Ways in | Desktop, SSH, VS Code, AI assistants | Web terminal, SSH, JupyterLab, VS Code | SSH, Jupyter | Notebook |
| When you stop | Paused: disk, apps and files kept | Volume kept, container disk cleared | Data kept on that host | Runtime ends, 12 hours at most |
| Getting the GPU back | Any free card of that size | Not held: a restart can get zero GPUs | Waits if the host's GPU is rented | Card not guaranteed |
| When you forget it | Pauses itself after 15 idle minutes | Runs until you stop it | Runs until you stop it | Times out when idle |
| Applications | ComfyUI, vLLM, Ollama in one click | Docker templates | Docker templates | Notebooks |
| Where | US East, Frankfurt | 31 regions | Wherever a host is | Not stated |
Swipe for Vast.ai and Google Colab →
An honest answer
When another one fits better
Hinode is not the cheapest hour of GPU. Here is who should rent elsewhere, and who should not.
Pick RunPod
when the hourly rate decides. The same L40S is $1.09 an hour there and $3.30 here, and it has 31 regions. You look after the Pod: stop it yourself, and keep what matters on a volume.
Hinode as a RunPod alternative →Pick Vast.ai
when you want the lowest price there is and can move when a host goes away. Prices change by the hour and each machine belongs to its own host.
Pick Google Colab
when a notebook is all you need and a session under twelve hours is enough. You do not choose the card.
Pick Hinode
when you come back to the same setup every day: a desktop, your apps and models where you left them, and a machine that stops billing when you walk away.
Create a workspace →Sources · read 2026-10-05
RunPod: runpod.io/pricing · docs.runpod.io/pods/storage/types · docs.runpod.io/pods/manage-pods · docs.runpod.io/pods/connect-to-a-pod · runpod.io/product/cloud-gpus
Vast.ai: its public offers list (verified hosts, one GPU, on demand) · docs.vast.ai/documentation/instances/pricing · docs.vast.ai/documentation/instances/manage-instances
Google Colab: research.google.com/colaboratory/faq.html
Hinode: the live price list, the same one the app and an invoice read.
FAQ
Comparing GPU clouds: questions
A Hinode machine is a whole computer of your own: a desktop, your apps and a disk that are still there after a pause. RunPod rents the same card for about a third of the price, and Vast.ai for less. When the hourly rate is what matters, rent there.
The other providers' prices were read from their own pages on 2026-10-05, and the page carries that date. Hinode's prices are the list an invoice uses. When a figure here differs from the provider's page, the provider's page is right.
Only for its disk: $0.15 per GB-month for what is written on it. The hourly charge stops, and a machine pauses itself after 15 idle minutes.
Yes. Every machine answers over SSH, so scp and rsync copy files in from another server or from your own computer. Put models in the shared folder and every machine you start reads them.
RunPod and Vast.ai start ComfyUI from a Docker template. On Hinode it installs in one click, your custom nodes and models are still there next time, and it opens in a tab on your own computer.
Set it up once. Find it there tomorrow.
Create a workspace, install your apps, and start a GPU when you need one.
