# DigitalOcean GPU Droplets

Official pricing: https://www.digitalocean.com/pricing/gpu-droplets  
Category: gpu-cloud · Isolation: vm

## Pricing regimes (raw)

- **MI325X, on-demand (×1)** (resource): $0/vCPU-h, $0/GiB-h [stock]
- **MI325X, on-demand (×8 node)** (resource): $0/vCPU-h, $0/GiB-h
- **MI300X, on-demand (×1)** (resource): $0/vCPU-h, $0/GiB-h
- **MI300X, on-demand (×8 node)** (resource): $0/vCPU-h, $0/GiB-h
- **H200, on-demand (×1)** (resource): $0/vCPU-h, $0/GiB-h
- **H200, on-demand (×8 node)** (resource): $0/vCPU-h, $0/GiB-h
- **H100, on-demand (×1)** (resource): $0/vCPU-h, $0/GiB-h
- **H100, on-demand (×8 node)** (resource): $0/vCPU-h, $0/GiB-h

## Features

Yes: snapshots, fork/clone, persistent disk, volumes, ≥24 h sessions, custom image (Docker or snapshot), start from your own snapshot, full VM (own kernel), Docker inside, root, SSH, public IPv4, raw TCP inbound, egress allowlist, open internet, static egress IP, private networking, SOC 2, HIPAA, SSO, GPU, Python, live fork (no pause), EU data residency, MCP server, snapshots on demand, gVisor or VM (no shared kernel), inbound access rules, firewall inside (nftables), extra volumes, shared volumes

No: memory snapshots, pause/resume, idle auto-stop, your own Docker/OCI image, browser, desktop GUI, computer-use API, browser + desktop control, anti-bot stealth, CAPTCHA solving, residential IPs, preinstalled agents, HTTPS preview URLs, self-hosting, BYOC, open source, arm64, Windows, macOS, Node.js, HTTP method/path egress rules, wake on request, live resize, memory fork, automatic snapshots, agent harness API, hosted agent API (their own agent), secret proxy, secret proxy for any API

Unknown: everything else. Evidence (source + quote) per feature: https://battleships.dev/data/providers/digitalocean-gpu.json → feature_evidence

## Caveats

- Per-second with 5 minute minimum; powered-off Droplets STILL bill until destroyed; GPU Droplets have no CPU Droplet monthly cap. August 1 2026 price revision. Bundled CPU/RAM and local disks; transfer allowance included.
- Any null count, CPU/RAM, billing increment or minimum is unverified, not unlimited/free.
- This is GPU-product research, not a claim that every feature of the entire vendor documentation was audited. Unverified features are null.
- GPU multi-count bundles, variant/region selection and fractional slices require a SKU-aware estimator; unsupported rows remain unpriced in strict cards.
- Modes with gpu=null are intentionally excluded, not free GPU modes. Never substitute 0 for a null rate. Flags alt indicate DIY/ML infrastructure, not the same thing as managed untrusted-code sandboxes.
- Unknown fee, storage or bandwidth means full monthly total may be unavailable. Rates exclude taxes.
- Billing minimum conflict: docs (2026-08-25) say 60 s or $0.01 minimum; pricing-page FAQ says 5-minute minimum. Monthly: GPU Droplets have NO 672-h cap ('If you use your GPU Droplet for more than the standard 672-hour monthly limit, you will still be charged for the time you use'); embedded catalog monthly = hourly x 744. Newer customers may be charged mid-cycle when creating GPU Droplets.
- Spot (public preview, Beta Service): no SLA, no support commitment, no service credits; 2-hour reclaim notice is best-effort ('may be reclaimed with less than two (2) hours' notice or ... with no notice'); boot and scratch disk data lost on reclaim; spot rate set independently of on-demand, locked at creation for the Droplet's life. Access 'limited to eligible accounts'.

## How this provider charges

As of 2026-09-28. GPU Droplets.

Per-second with 5 minute minimum; powered-off Droplets STILL bill until destroyed; GPU Droplets have no CPU Droplet monthly cap. August 1 2026 price revision. Bundled CPU/RAM and local disks; transfer allowance included.

| Regime | When applicable | Billing | Numbers ($/GPU-hour unless slice) | Source |
|---|---|---|---|---|
| on-demand  | MI325X ; count [1]; Region/availability must be checked at launch. | per-second; min 300s | 3.8; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| on-demand  | MI325X ; count [8]; Region/availability must be checked at launch. | per-second; min 300s | 3.8; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| on-demand  | MI300X ; count [1]; Region/availability must be checked at launch. | per-second; min 300s | 2.59; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| on-demand  | MI300X ; count [8]; Region/availability must be checked at launch. | per-second; min 300s | 2.59; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| on-demand  | H200 ; count [1]; Region/availability must be checked at launch. | per-second; min 300s | 4.47; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| on-demand  | H200 ; count [8]; Region/availability must be checked at launch. | per-second; min 300s | 4.47; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| on-demand  | H100 ; count [1]; Region/availability must be checked at launch. | per-second; min 300s | 4.41; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| on-demand  | H100 ; count [8]; Region/availability must be checked at launch. | per-second; min 300s | 4.41; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| on-demand  | RTX-4000-ADA ; count [1]; Region/availability must be checked at launch. | per-second; min 300s | 0.76; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| on-demand  | RTX-6000-ADA ; count [1]; Region/availability must be checked at launch. | per-second; min 300s | 1.57; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| on-demand  | L40S ; count [1]; Region/availability must be checked at launch. | per-second; min 300s | 1.57; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| reserved  | B300 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | 7.94; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| reserved  | H200 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | 3.4; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| reserved  | H100 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | 3.26; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| reserved  | MI350X ; count None; Region/availability must be checked at launch. | per-second; min unknowns | 4.76; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| reserved  | MI325X ; count None; Region/availability must be checked at launch. | per-second; min unknowns | 2.88; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| reserved  | MI300X ; count None; Region/availability must be checked at launch. | per-second; min unknowns | 1.91; list; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| spot  | B300 ; count None; Region/availability must be checked at launch. | per-second; min 300s | 8; dynamic-snapshot; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| spot  | MI355X ; count None; Region/availability must be checked at launch. | per-second; min 300s | 2.97; dynamic-snapshot; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |
| spot  | MI350X ; count None; Region/availability must be checked at launch. | per-second; min 300s | 2.46; dynamic-snapshot; per_gpu_hour | [source](https://www.digitalocean.com/pricing/gpu-droplets) |

## Gotchas
- Per-second with 5 minute minimum; powered-off Droplets STILL bill until destroyed; GPU Droplets have no CPU Droplet monthly cap. August 1 2026 price revision. Bundled CPU/RAM and local disks; transfer allowance included.
- Any null count, CPU/RAM, billing increment or minimum is unverified, not unlimited/free.
- This is GPU-product research, not a claim that every feature of the entire vendor documentation was audited. Unverified features are null.
- GPU multi-count bundles, variant/region selection and fractional slices require a SKU-aware estimator; unsupported rows remain unpriced in strict cards.

## Worked example (required protocol workload)
4 vCPU / 8 GiB, 50 concurrent × 8 hours/day × 22 days = **8,800 instance-hours/month**. 30% CPU utilization does not reduce allocated GPU uptime. 50 GiB snapshots and 100 GiB egress are additional. The requested CPU-only workload has no GPU type/count, so its full GPU-provider total is **null**, not a fictitious CPU equivalent.

Explicit GPU sensitivity example: add **1 × MI325X** to each of those 50 workers at the MI325X quoted shape. Compute component = 8,800 × $3.8 = **$33,440.0000**. CPU/RAM are included in that SKU (not independently resizable).
This is compute-only, before platform fees, storage, egress, setup, minimum rounding, cold-start/idle tails and capacity quotas. 50 concurrent GPUs are not promised by a unit rate.

## Evidence scope
Sources read are listed in the provider/features JSON. Raw pricing pages, relevant docs and API payloads are retained under `../raw/`. All unknown feature toggles remain null.