Paperspace Core
Server virtualization software
- Features
- Ease of use
- Ease of management
- Quality of support
- Affordability
- Market presence
Take the quiz to check if Paperspace Core and its alternatives fit your requirements.
Pay-as-you-go
Free Trial unavailableFree version unavailable
Small
Medium
Large
- Media and communications
- Education and training
- Information technology and software
What is Paperspace Core
Paperspace Core is a cloud infrastructure platform for provisioning and managing virtual machines, including GPU-backed instances, through a web console and API. It is used by teams that need on-demand compute for development environments, rendering, and machine learning workloads without operating their own virtualization stack. The service focuses on rapid VM deployment, snapshotting, and remote access workflows rather than on-prem hypervisor management. Paperspace Core operates as a managed service under DigitalOcean following DigitalOcean’s acquisition of Paperspace.
GPU VM provisioning options
Paperspace Core offers virtual machines with GPU configurations suited to compute-intensive workloads such as model training and graphics rendering. This reduces the need to assemble and maintain dedicated GPU servers. For teams comparing general-purpose VM services, the GPU-oriented catalog and workflows are a practical differentiator for specific use cases.
API and automation support
The platform provides an API for programmatic provisioning and lifecycle management of instances. This supports repeatable environments, CI/CD-driven infrastructure changes, and integration with internal tooling. Automation capabilities help teams scale usage up or down without relying solely on a manual console workflow.
Snapshots and instance lifecycle tools
Paperspace Core includes common VM management functions such as snapshots, templates/images, and instance start/stop controls. These features support fast environment cloning and rollback for development and experimentation. They also help reduce setup time when teams need multiple similar machines for short-lived projects.
Cloud-only, not on-prem
Paperspace Core is delivered as a managed cloud service rather than software you deploy in your own data center. Organizations that require on-premises virtualization, air-gapped environments, or strict data residency controls may find the deployment model limiting. This can also constrain customization of the underlying hypervisor and networking stack compared with self-managed virtualization platforms.
Ecosystem breadth varies
Compared with larger cloud compute ecosystems, the breadth of adjacent managed services (for example, tightly integrated networking, identity, and platform services) may be more limited depending on the workload. Teams may need to assemble additional components outside the platform for end-to-end production architectures. This can increase integration effort for complex enterprise deployments.
Service changes after acquisition
Because Paperspace is owned by DigitalOcean, product packaging, availability, and roadmap can change as services are consolidated. Buyers may need to validate current feature parity, regional availability, and support processes at the time of purchase. This adds due diligence overhead for long-term standardization decisions.
Plan & Pricing
Pricing model: Pay-as-you-go (on-demand instances billed by the hour/second; monthly reserved pricing also shown). Free tier/trial: No permanent free Core compute tier or time-limited Core trial is shown on the Core pricing page (account creation is free but Core instances are billed).
Example costs (official Paperspace Core pricing page):
-
Dedicated GPU (hourly and monthly examples):
- M4000 — $0.45 per hour; $268 per month — 8 GB GPU; 30 GB RAM; 8 vCPU.
- P4000 — $0.51 per hour; $303 per month — 8 GB GPU; 30 GB RAM; 8 vCPU.
- P5000 — $0.78 per hour; $461 per month — 16 GB GPU; 30 GB RAM; 8 vCPU.
- P6000 — $1.10 per hour; $647 per month — 24 GB GPU; 30 GB RAM; 8 vCPU.
- V100 — $2.30 per hour; $1,348 per month — 16 GB GPU; 30 GB RAM; 8 vCPU.
- RTX4000 — $0.56 per hour; $337 per month — NVIDIA RTX4000; 30 GB RAM; 8 vCPU.
- RTX5000 — $0.82 per hour; $484 per month — NVIDIA RTX5000; 30 GB RAM; 8 vCPU.
- A4000 — $0.76 per hour; $488 per month — NVIDIA A4000; 44–45 GB RAM; 8 vCPU.
- A5000 — $1.38 per hour; $891 per month — NVIDIA A5000; 44 GB RAM; 12 vCPU.
- A6000 — $1.89 per hour; $1,219 per month — NVIDIA A6000; 90 GB RAM; 12 vCPU.
- A100-80G — $1.15 per hour (multi-GPU & 8x options listed).
- H100 / HGX H100 — listed as available (hourly and/or reserve) — H100 hourly and reserve pricing shown on the official page.
-
CPU Servers (Servers / Desktops):
- C1 — $6 per month (1 vCPU, 512 MB RAM) or $0.0045/hr equivalent shown for low-tier CPU types.
- C2 — $9 per month (1 vCPU, 1 GB RAM) ($0.009/hr shown).
- C3 — $14 per month (2 vCPU, 2 GB RAM) ($0.018/hr shown).
- C4 — $25 per month (2 vCPU, 4 GB RAM) ($0.04/hr shown).
- C5 — $48 per month (4 vCPU, 8 GB RAM) ($0.08/hr shown).
- C6 — $92 per month (8 vCPU, 16 GB RAM) ($0.16/hr shown).
- C7 — $158 per month (12 vCPU, 30 GB RAM) ($0.30/hr shown).
- C8 — $278 per month (16 vCPU, 60 GB RAM) ($0.60/hr shown).
- C9 — $408 per month (24 vCPU, 120 GB RAM) ($0.90/hr shown).
- C10 — $788 per month (32 vCPU, 244 GB RAM) ($1.60/hr shown).
Notes & limits:
- Official site indicates instances are billed per second (on Core) and shows both hourly and monthly prices for many instance types.
- Multi-GPU variants (2x/4x/8x) are available (documentation/spec note on the pricing page).
- Some very large GPUs (A100, H100) show special hourly/availability notes and a "Reserve Now" flow for certain SKUs.
(Prices and specs taken directly from Paperspace's official pricing pages for Core / GPU Cloud and the Core product page.)
Seller details
DigitalOcean Holdings, Inc.
New York, NY, USA
2011
Public
https://www.digitalocean.com/
https://x.com/digitalocean
https://www.linkedin.com/company/digitalocean/
