GPU Droplets
thecontinuitylayer.com/tools/gpu-hosting/digitalocean-gpu-dropletsDevelopers already on DigitalOcean who want a GPU server next to their app
GPU cloud · DigitalOcean
GPU Droplets are DigitalOcean's GPU virtual machines, offered in single-GPU and 8-GPU sizes with both NVIDIA and AMD cards. They sit in the same console as DigitalOcean's regular servers, databases and storage. Paperspace, which DigitalOcean owns, now runs on the same platform.
For open models, the smaller cards suit 7B to 30B-class models and the H100, H200 and MI300X handle larger ones. Setup is simple, but the big GPUs cost more per hour than at specialist GPU clouds, and 12-month reservations are needed for the lower rates.
Specs
| Price per hour | RTX 4000 Ada $0.76/hr; L40S $1.57/hr; H100 $4.41/hr per GPU on demand |
|---|---|
| GPUs | NVIDIA RTX 4000 Ada, L40S, H100, H200 and AMD MI300X |
| RTX 4000 Ada | $0.76 per GPU-hour |
| L40S | $1.57 per GPU-hour |
| H100 | $4.41 per GPU-hour on demand, $3.26 reserved 12 months |
| H200 | $4.47 per GPU-hour on demand |
| AMD MI300X | $2.59 per GPU-hour on demand |
| Billing | Per second, with a short minimum charge |
Good for
- Teams already using DigitalOcean
- Mid-size open models on a single card
- Trying AMD MI300X
Not for
- The lowest H100 price
- Per-request serverless billing
See also
- Hyperbolic · Developers who want on-demand data-center GPUs without a contract
- Modal · Python developers who want to serve an open model as code, paying only while it runs
- Nebius AI Cloud · Teams that want a full-featured AI cloud with self-service access to recent NVIDIA GPUs
We last reviewed GPU Droplets on 31 October 2026. Prices and specs change; check digitalocean.com before you buy or sign up. Figures we could not confirm on the vendor's own page read “not published” or “varies”.
Vendor source: https://docs.digitalocean.com/products/droplets/details/pricing/