Packet.ai GPU Pricing
An on-demand GPU cloud operated by hosted.ai with dynamic and dedicated GPU instances plus reserved multi-node clusters.
Lowest Price
NVIDIA GeForce RTX 4090
Live Inventory
Real-time GPU availability on Packet.ai
| Provider | GPU Model | VRAM | Host Specs | Region | Price | Status | Action | |
|---|---|---|---|---|---|---|---|---|
Packet.ai | 2×NVIDIA GeForce RTX 4090 24GB VRAM | 24GB | 32 vCPU 128GB RAM 500GB Storage | 🌍global | $0.39/GPU/hr $0.77/hr total (2×) | Sold Out | ||
Packet.ai | NVIDIA RTX PRO 6000 Blackwell 96GB VRAM | 96GB | 6 vCPU 96GB RAM 150GB Storage | Santa Clara, California | $0.66/GPU/hr | Available | ||
Packet.ai | NVIDIA RTX PRO 6000 Blackwell 96GB VRAM | 96GB | 6 vCPU 96GB RAM 150GB Storage | Santa Clara, California | $0.66/GPU/hr | Sold Out | ||
Packet.ai | NVIDIA L40S 48GB VRAM | 48GB | 8 vCPU 64GB RAM 50GB Storage | 🌍global | $0.92/GPU/hr | Sold Out | ||
Packet.ai | NVIDIA L40S 48GB VRAM | 48GB | 8 vCPU 64GB RAM 50GB Storage | Reading | $0.92/GPU/hr | Sold Out |
About Packet.ai
Best For
Unique Features
- Authenticated launch API reports capacity by pool and region
- Dynamic and dedicated allocation modes
- Direct GPU-prefilled signup links
- Open-source CLI for provisioning and SSH access
Limitations
- Availability varies by GPU count and region
- Some GPU families are offered only through reserved clusters or waitlists
- Public catalog and live capacity must be checked separately
Platform Features
| Feature | Available |
|---|---|
| SSH Access | |
| Jupyter Notebooks | |
| Web Terminal | |
| API Access | |
| Kubernetes | |
| Containers |
| Feature | Available |
|---|---|
| Billing Increment | per-hour |
| Spot Instances | |
| Reserved Instances | |
| Prepaid Credits |
| Certification | Certified |
|---|---|
| SOC 2 | |
| HIPAA | |
| GDPR | |
| ISO 27001 |
| Feature | Available |
|---|---|
| SLA Guarantee | |
| Enterprise Support | |
| Discord Community |