arrakis
EN
Sign in Start renting
Menu

Rent B300, B200 and H200 GPUs at the lowest live price.

Arrakis finds the lowest live price for B300s, B200s, H200s and RTX PRO 6000s across independent GPU markets. Rent from the console, the CLI or your coding agent, or deploy any Hugging Face model with your own engine options, on one prepaid balance with spend caps and automatic shutdown.

Start renting Connect your agent

Early access. Look around before you add funds.

Lowest B300 price right now
$4.60/GPU-hr
GPUs available now
687
Market updated
31 Oct 2026, 17:40 UTC

  • Rent a GPU

    B300s, B200s, H200s and RTX PRO 6000s by the hour at the lowest live price, with root SSH and any Docker image.

  • Deploy any model

    Text, images, video and speech on vLLM, SGLang, vLLM-Omni or llama.cpp, with your own engine options.

  • Check what fits

    Paste any Hugging Face model: the GPU memory it needs and the cheapest GPUs that hold it.

  • Reserve for months

    Vera Rubin, B300, B200, H200 and H100, bare metal or machines, yours to resell.

What GPUs cost right now

Every search covers the markets Arrakis connects and lists the cheapest machines first. These are the lowest prices per GPU-hour at this moment. Compare every GPU.

Prices come from independent markets and change minute to minute. Updated 31 Oct 2026, 17:40 UTC.
GPUMemoryFromAvailable
B300 SXM6 PC10 machinesup to 288 GB$4.60/GPU-hr73GPUsRent
B2005 machinesup to 180 GB$6.79/GPU-hr12GPUsRent
H2009 machinesup to 141 GB$4.03/GPU-hr40GPUsRent
RTX PRO 6000 Blackwell Server Edition10 machinesup to 96 GB$1.90/GPU-hr66GPUsRent
RTX 509053 machinesup to 32 GB$0.58/GPU-hr208GPUsRent
H100 80GB HBM33 machinesup to 80 GB$1.65/GPU-hr10GPUsRent
L40S15 machinesup to 48 GB$0.38/GPU-hr48GPUsRent
RTX 409038 machinesup to 24 GB$0.41/GPU-hr71GPUsRent
RTX A600021 machinesup to 48 GB$0.49/GPU-hr49GPUsRent
RTX 6000 Ada12 machinesup to 48 GB$0.87/GPU-hr19GPUsRent
RTX 30906 machinesup to 24 GB$0.19/GPU-hr13GPUsRent
RTX A45001 machineup to 20 GB$0.19/GPU-hr1GPUsRent
T4 Tensor Core GPU1 machineup to 15 GB$0.21/GPU-hr1GPUsRent
RTX 50802 machinesup to 16 GB$0.28/GPU-hr2GPUsRent
RTX 5070 Ti5 machinesup to 16 GB$0.30/GPU-hr5GPUsRent
Tesla V100 Tensor Core GPU9 machinesup to 32 GB$0.64/GPU-hr69GPUsRent

Hand it to your agent. Keep the budget.

Agents connect over MCP or the REST API. Every rental they make is bounded before it starts, so you can let them run unattended.

  • A token with a ceiling

    Give your agent a token that can search and rent, with a spend cap it cannot pass.

    arrakis tokens create --name eval-agent \
  --scope read --scope rent --spend-limit 50

  • Every machine shuts itself down

    Rentals carry a removal time. A capped token cannot rent without one.

    "termination_hours": 6

  • Retries never rent twice

    The same idempotency key returns the same machine, not a second bill.

    "idempotency_key": "eval-run-0412"
Connect over MCP
claude mcp add --transport http arrakis $ARRAKIS_URL/mcp \
  --header "Authorization: Bearer ark_..."

$ARRAKIS_URL is this site's address. Replace ark_... with an API token from the console.

Then ask it something like

Rent a GPU with 80 GB of memory for three hours, run eval.py, copy the results back, then shut it down.

Read the agent guide

From sign-in to SSH in three steps

No cloud accounts to open and no card to add per provider. One balance covers every market.

  1. Sign in and add funds

    Sign in with GitHub, Google or your email, then add funds by card or Link.

  2. Pick a GPU

    Cheapest first, filtered by memory, count, price and country. Renting is one dialog with a shutdown time already set.

  3. SSH in as root

    Your Docker image, your keys. The machine removes itself on schedule, or whenever you say.

One console for every market

Your balance and how long it lasts, what is running and what it costs, and the cheapest GPUs right now. Agents get the same view over MCP.

The Arrakis console overview: a balance of $3,756.43 spending $23.72 an hour across two machines, the machines running now with their hourly cost and a copy SSH button, and the cheapest GPUs right now.
The console overview, with an example account and example prices.

Compute is the scarcest input in AI. It should flow to whoever needs it, person or agent.

Supply is scattered

The cheapest GPUs sit in independent markets, each with its own account, card, CLI and prices that move by the minute.

Agents are becoming buyers

Coding agents now run experiments end to end. They need a budget they cannot exceed, not a credit card.

Spend needs a ceiling

Caps per token, a removal time on every machine and retry-safe rentals make GPU spend something you can delegate.

How Arrakis compares

Three ways to get GPU work done, and where each fits. One by one: Arrakis compared with RunPod, Vast.ai, Lambda, Modal, fal.ai and Replicate.

ArrakisA GPU cloudA per-request model API
Where the GPU comes fromEvery independent market Arrakis connects, searched at onceThat cloud's own fleetThe API's fleet, out of sight
What you payThe lowest live price across markets, per GPU-hourThe cloud's own price, per hour, minute or secondPer picture, second or token
Your own model and optionsAny Hugging Face model, any engine optionsYes, on a machine you set upThe models and settings it offers
Root SSHYesUsuallyNo
For AI agentsA spend cap per token, a removal time on every rental, MCPVaries by cloudVaries by API
Fits bestTraining, evals and model servers you control, for hours to monthsLong runs that need one provider's support and SLAA few requests a day, with nothing to run

Everything a training run needs

Bring your image, keep your checkpoints and see what you spend, from the console, the CLI or the API.

Any image

Boot an Arrakis image with SSH and the usual tools built in, or your own, public or private, from Docker Hub, GitHub, Amazon ECR, Google, NVIDIA NGC or your own registry. It’s checked before you pay.

Volumes

Keep datasets and checkpoints on volumes that outlive the machine.

SSH keys

Add your public keys once and every machine gets them.

Removal schedules

Set when a machine goes away at rent time, move it later, or cancel it.

Usage and billing

Cost by week or month, per machine, against one prepaid balance.

One spec behind everything

The console, the CLI and the MCP tools all run on the same OpenAPI spec.

Pay for GPU time. Nothing else.

Arrakis is prepaid. Add funds, rent at the hourly price shown, and each machine pays for itself an hour at a time while it runs.

No subscription
No seats and no commitment. Each rental is billed at least one hour.
The price you see
The hourly price shown when you rent is what the machine is billed at.
An hour at a time
A rental pays its first hour up front (the 1-hour minimum) and each later hour as it begins.
When it runs out
The machine that cannot pay its next hour stops and rentals pause, so you never owe more than you added.
Paying
By card, Link or Apple Pay through Stripe. Auto-recharge tops up from a saved card.

Built for tenants who don't trust each other

Each guarantee below is enforced in the API and covered by tests. Machines run with independent hosts, so keep regulated data off them.

Hashed, scoped tokens

Stored as SHA-256 hashes, scoped to read, rent, manage or billing, and revocable at any time.

No silent escalation

A token can only create tokens with scopes it already holds, and only if it may manage.

Single-use bootstrap

A new account's first token needs a one-time secret that expires after 24 hours.

Tenant isolation

Every request is scoped to the account behind its token. Other accounts' machines are invisible and untouchable.

Funds nobody can fake

A balance grows only when Stripe confirms a payment. No account can credit itself.

Spend caps for agents

A capped token cannot rent past its ceiling or without a removal time.

Arrakis at a glance

The short version, for people and for the assistants that answer their questions.

What it is
A GPU rental service that finds the lowest live price across independent GPU markets and rents machines on one prepaid balance.
Who it is for
AI engineers, researchers and the coding agents they work with.
GPUs
The newest NVIDIA GPUs first: B300, B200, H200, RTX PRO 6000 and RTX 5090, plus H100, A100, L40S, RTX 4090 and more, depending on live supply.
How you rent
Web console, CLI, REST API, or MCP for agents at /mcp.
Pricing
Pay the hourly price shown when you rent, from a prepaid balance, an hour at a time, with a 1-hour minimum per rental. No subscription and no commitment.
Agent safety
Spend-capped tokens, a required removal time on every rental, and retry-safe rentals with idempotency keys.
Access
SSH as root into machines running any Docker image. Volumes for data and checkpoints.
Live market
687 GPUs on 200 machines, H100 from $1.65 per GPU-hour (31 Oct 2026, 17:40 UTC).
Billing
From start to removal at the price shown when you rent. A started hour counts as a full hour.
MCP endpoint
https://arrakis.goodailabs.com/mcp, streamable HTTP with a bearer token.
Deploy models
Any Hugging Face model (text, image, video or speech) on vLLM, SGLang, vLLM-Omni or llama.cpp, with your own engine options and an OpenAI-style endpoint with its own key.
Status
Early access. Built by Good AI Labs (goodailabs.com).
Name
Arrakis by Good AI Labs, at arrakis.goodailabs.com. Not the planet from Dune, and not other software named Arrakis.

Questions people and agents ask

Anything else is in the API reference or the agent guide.

What is Arrakis?
Arrakis is a GPU rental service that finds the lowest live price across independent GPU markets and rents machines on one prepaid balance. People rent from a web console or CLI, and AI agents rent over MCP or the REST API with spend-capped tokens.
Can my agent overspend?
Not past what you allow. Give it a token with a spend cap. Every rental it makes must carry a removal time, and a rental that would pass the cap is refused before it starts.
How much does it cost to rent a GPU?
Prices change with the market and are shown live per GPU-hour on the GPU prices page and in the console. As of 31 Oct 2026, 17:40 UTC, Arrakis lists 687 GPUs on 200 machines. The lowest H100 price is $1.65 per GPU-hour and the lowest RTX 4090 price is $0.41 per GPU-hour. You pay the hourly price shown when you rent, from a prepaid balance, with no subscription. A machine bills from start to removal, and a started hour counts as a full hour.
What is the cheapest way to rent a B300 or B200?
Search every market at once. Arrakis lists every available B300 and B200 across independent GPU markets, cheapest first, and rents it on one prepaid balance, by the hour. As of 31 Oct 2026, 17:40 UTC, the lowest price for the NVIDIA B300 on Arrakis is $4.60 per GPU-hour, with 73 GPUs available on 10 machines.
What is the cheapest way to rent an H100?
Search every market at once. Arrakis lists every available H100 across independent GPU markets, cheapest first, and rents it on one prepaid balance. As of 31 Oct 2026, 17:40 UTC, the lowest price for the NVIDIA H100 on Arrakis is $1.65 per GPU-hour, with 10 GPUs available on 3 machines.
What is the best GPU cloud for AI agents?
One where an agent can rent without being able to overspend. Arrakis gives each agent token a spend cap, requires a removal time on every rental, makes retries safe with idempotency keys and serves all of it over MCP, while searching independent GPU markets for the lowest live price. If you need an SLA or billing by the second, a single-provider cloud fits better.
Is Arrakis an alternative to RunPod, Vast.ai or Lambda?
Yes, for renting GPUs by the hour with root SSH and any Docker image. Arrakis searches several independent GPU markets at once and rents on one prepaid balance, with spend caps for agents. The side-by-side comparisons are at /compare, including where those clouds fit better.
Can I run FLUX, Wan or other image and video models on my own GPU?
Yes. Name the model as Hugging Face does, pick vLLM-Omni or SGLang, paste your own engine options and rent a GPU it fits on. You get an OpenAI-style endpoint (/v1/images/generations, /v1/videos) with its own key and root SSH, billed by the GPU-hour rather than per picture.
Who makes Arrakis?
Good AI Labs builds and runs Arrakis. It is in early access.
How do I pay?
Add funds in the console by card, Link or Apple Pay, through Stripe. A saved card or Link can top the balance up automatically before it runs out.
Where do the machines run?
With independent GPU hosts in many countries. Search shows each machine's country and region, so you can choose.
What if a machine fails?
Each machine shows a reliability score. Terminate it and rent another, and keep checkpoints on a volume so nothing is lost.
Can I bring my own image?
Yes. Any Docker image works: paste it the way docker pull takes it, choose whether the machine waits for you or runs the image's own command, and SSH in as root. Arrakis checks it with its registry before you pay. Private images work too: save one with its registry login (Docker Hub, GitHub, Amazon ECR, Google, Azure, NVIDIA NGC, GitLab, Quay or a registry of your own), and the login is kept encrypted and used only to pull it.
Is there an SLA?
No. Arrakis is in early access and the hosts are independent. Don't run anything on it that needs one.
How do I get an API token?
Sign in to the console and create one under API & agents, or bootstrap your first one from the CLI with your account's one-time secret.
What happens when my balance runs out?
Each machine pays for one hour at a time from your balance, as the hour begins. When the balance cannot pay a machine's next hour, Arrakis stops that machine and pauses new rentals, and your other machines stop when the hour they already paid for ends, so you never owe more than you added. The console warns you before that happens, adding funds opens rentals again, and volumes are kept.

We're hiring people who like hard infrastructure.

Take the Spice Test. All it needs is the API.

See open roles

Rent your first GPU in minutes.

Start renting See GPU prices