Support

The quickest way to get help is the Support item in the dashboard sidebar. It opens a Contact Support form with a category, a subject and a description, and it reaches the same team as email to support@deploygpu.ai. There is no phone line, live chat or Slack channel, and no response time is promised.

​

What to include

For a problem with a deployment, include its deployment ID, the GPU and provider, roughly when it happened, and any failure code shown in the dashboard or the email you received. For a payment, include the date and amount. Never send card numbers or private keys.

​

Common issues

My deployment is stuck in provisioning

A GPU VM usually reaches Running in 1 to 3 minutes, and an app template in about 6 to 8, because the platform waits for the app itself to answer. If a launch stays in provisioning well past that, stop it and try another offer: stopping before Running refunds the whole reserved hour, and a launch that fails is refunded automatically.

I cannot SSH into my instance

Copy the exact SSH command from the deployment, since the user differs by provider. The connect over SSH guide covers keys and the usual errors, and the site has fixes for Permission denied (publickey) and Connection refused.

My deployment stopped or disappeared

There are 3 usual causes. When your balance cannot cover the next hour, the instance switches to per-minute billing until the balance reaches zero and is then stopped, with emails at about 2 hours left, when per-minute billing starts and when it stops. If the provider loses the machine, a health check that runs every 5 minutes ends the deployment and refunds the unused time, without an email, and it no longer appears in the dashboard list. On QuantaCloud and Massed Compute, a VM shut down from inside counts as lost the same way. Write to support with the deployment ID if none of these fits.

I lost my SSH private key

The private half of a key generated by DeployGPU can only be downloaded once. Generate or add a new key in SSH Keys and launch a new deployment with it selected. Copy any data you need off the old instance first if you can still reach it.

My volume is in Error state

Delete the volume and create a new one. If it happens again in the same region, try another region, and tell support which provider and region it was.

A payment failed

Check the card details and that your bank is not blocking the payment, then try the top-up again from Billing. Card top-ups go through Stripe and must be between $5 and$25,000.

​

Data and privacy

• Instance data: when a deployment stops, its disk is deleted. Use volumes on Verda or Hyperstack, or copy files off before you stop.
• Volume data: stays in the region where it was created until you delete the volume, or until your balance cannot pay for it, in which case it is deleted permanently without an email.
• Account data: your profile, billing history and SSH key metadata. See Settings to delete your account.

​

Service availability

DeployGPU has no SLA or uptime guarantee. GPU stock depends on what each provider has free at that moment, so if a GPU type or region is out of stock, try another configuration in the marketplace or check again later.