Paperspace
Paperspace, now part of DigitalOcean, provides an accelerated cloud computing platform specifically designed for AI and Machine Learning workloads.
Our verdict
Who it's for
Build, train, and deploy AI/ML models on accelerated cloud GPUs with simplicity and scalability.
Who should look elsewhere
Specific instance types and their availability may vary.
Pros
- +Significantly reduces compute costs compared to major public clouds or self-hosting.
- +Simplifies AI/ML infrastructure management, allowing focus on model development.
- +Offers flexible, on-demand scaling with no long-term commitments.
- +Provides a comprehensive platform for the entire ML lifecycle, from development to deployment.
- +Includes collaboration features for team-based projects.
Cons
- −Specific instance types and their availability may vary.
- −Free tier has limitations on storage and auto-shutdown duration.
Key features
- ✓NVIDIA H100 GPU access for AI/ML workloads
- ✓ML platform for building, training, and deploying models (Gradient)
- ✓Fully-managed cloud GPU platform (CORE)
- ✓1-click hosted Jupyter Notebooks
- ✓End-to-end MLOps capabilities
- ✓Low-cost GPUs with per-second billing
- ✓On-demand pricing with no commitments
- ✓Instant provisioning of instances
Pricing
| Plan | Price | Billing | Notes |
|---|---|---|---|
| Free | $0 | month | FREE GPU; Public projects; Auto-shutdown (12 hour limit) |
| Pro | $8 | month | Private projects; 15GB storage; Auto-shutdown (configurable) |
| Growth | $39 | month | Private projects; 50GB storage; Auto-shutdown (configurable) |
| T0 Small Teams | $0 | month | Low Instance Types; Public Notebooks; 10GB Persistent Storage included |
| T1 Mid-Size Teams | $12 | month | Low - Mid Instance Types; Private Notebooks; 500GB Persistent Storage included |
| T2 Large Teams | — | month | Low - High Instance Types; Private Notebooks; Scalable Storage |
| Enterprise | — | month | Host Gradient on your own infrastructure; Run without limits; Custom plans |