Refund policy

1. What is the full rule?

If an instance fails to start, or its GPU does not work and support cannot fix it, the hours it used go back to your balance. Unused GPU-hours and balance never expire. Top-ups are not paid out unless the law requires it; the part added only to reach the minimum top-up comes back on request within 30 days of payment, in USDT or USDC, as long as it is unspent.

A person reviews each request. Once it is approved, the hours appear on your balance and can pay for any GPU straight away.

2. Case 1: the instance never starts

You start an H100 instance. It never reaches a running state, or it runs but no SSH connection ever succeeds. You write to support with the instance ID and the error you see.

Support looks at it first. If it cannot get the instance working, every minute charged for that instance goes back to your balance, and you can start a new one when a GPU is free.

3. Case 2: the GPU does not work

Your A100 instance starts, but nvidia-smi shows no GPU, or the GPU reports memory errors and your jobs crash on it. You write to support with the instance ID and the error output.

Support tries to fix it. If it cannot, the hours the instance used go back to your balance.

A GPU that works but runs your job slower than you hoped is not faulty. If a model does not fit or runs slowly, the VRAM guide helps you choose a better GPU for the next run.

4. Case 3: the minimum top-up was more than you needed

You want the 10-hour RTX 4090 block at $2.80. A top-up is at least $25, so you send the minimum and buy the block from it. The difference stays on your balance.

If you leave that difference unspent, ask for it within the time limit in the rule above and it comes back to you in USDT or USDC. If you spend part of it on more GPU time, only the unspent part can come back.

5. What is not refunded?

• GPU time used by a working instance, including time it ran idle before you stopped it. Instances are billed per minute at the hourly rate while they run, from the GPU-hours you bought for that GPU. Stopping an instance stops the charge at once.
• Hours used by an instance stopped for breaking the acceptable use policy.
• The rest of a balance, and unused GPU-hours. They are not paid out unless the law requires it, but they never expire: balance pays for any GPU, and hours stay available for their GPU model.
• Lost data or interrupted jobs, such as files on a deleted instance or a run stopped because the balance reached zero.
• Waiting for a busy GPU. Nothing is charged until an instance starts, so there is nothing to refund.

6. What about my rights under consumer law?

They apply on top of this policy. If the law of the country where you live gives you a right to withdraw from a purchase or to get money back, that right is respected, and nothing on this page reduces it.

7. How do I ask for a refund?

Email [email protected] and say which case applies. Include:

• For case 1 or 2: the instance ID, the GPU model, the time in UTC and the error output, for example from nvidia-smi or your SSH client.
• For case 3: the transaction ID of the top-up, and whether you want USDT or USDC and on which network. Support confirms the receiving address with you before anything is sent.

Nobody from our team will ever ask for your account number, your SSH private key or your API keys, by email or in a chat. A person answers support emails within 12 hours, seven days a week, and helps you start, size and stop your GPU instances.

8. Why do refunds go to the balance rather than to my wallet?

Crypto transfers have no chargeback, and each new transfer costs a network fee. Hours on the balance cost nothing to move and can pay for any GPU the same minute.

Coins leave our address in one case only, the minimum top-up difference of case 3, and wherever a law makes a payout compulsory.