AI Inference
Akamai Preferred Partner
Akamai Preferred Partner
The right price is not just competitive but predictable. GPUs start at $0.52/hour, and the pricing model is designed for better operational planning.
Balanced price and performance for AI image generation, small language models, video transcoding, and rendering. Includes 2x encoding, 2x decoding, and 1x AV1 encode/decode engines per card.
| Plan | GPU Cards | vCPUs | RAM | Storage | Hourly | Monthly |
|---|---|---|---|---|---|---|
| RTX 4000 Ada x1 Small | 1 | 4 | 16 GB | 512 GB | $0.52/hr | $350/mo |
| RTX 4000 Ada x1 Medium | 1 | 8 | 32 GB | 512 GB | $0.67/hr | $446/mo |
| RTX 4000 Ada x1 Large | 1 | 16 | 64 GB | 512 GB | $0.96/hr | $638/mo |
| RTX 4000 Ada x1 X-Large | 1 | 32 | 128 GB | 512 GB | $1.53/hr | $1,022/mo |
| RTX 4000 Ada x2 Small | 2 | 8 | 32 GB | 1 TB | $1.05/hr | $700/mo |
| RTX 4000 Ada x2 Medium | 2 | 16 | 64 GB | 1 TB | $1.34/hr | $892/mo |
| RTX 4000 Ada x2 Medium High Storage | 2 | 16 | 64 GB | 2 TB | $1.49/hr | $992/mo |
| RTX 4000 Ada x4 Small | 4 | 32 | 128 GB | 2 TB | $2.96/hr | $1,976/mo |
| RTX 4000 Ada x4 Medium | 4 | 48 | 196 GB | 2 TB | $3.57/hr | $2,384/mo |
Pricing matches Akamai Cloud’s published rates, verified June 2026. Egress as low as $0.005/GB in most regions — up to 90% lower than hyperscalers.
Proven professional GPUs for certified CAD, design visualization, and graphics workflows.
| Plan | GPU Cards | vCPUs | RAM | Storage | Hourly | Monthly |
|---|---|---|---|---|---|---|
| Dedicated 32GB + RTX6000 x1 | 1 | 8 | 32 GB | 640 GB | $1.50/hr | $1,000/mo |
| Dedicated 64GB + RTX6000 x2 | 2 | 16 | 64 GB | 1.25 TB | $3.00/hr | $2,000/mo |
| Dedicated 96GB + RTX6000 x3 | 3 | 20 | 96 GB | 1.9 TB | $4.50/hr | $3,000/mo |
| Dedicated 128GB + RTX6000 x4 | 4 | 24 | 128 GB | 2.5 TB | $6.00/hr | $4,000/mo |
Akamai’s newest server-class GPU for large-scale AI inference, fine-tuning, agentic AI, and 8K media pipelines — available in 1-card, 2-card, and 4-card plans. Blackwell capacity is in limited availability and provisioned by request.
NVIDIA RTX PRO™ 6000 Blackwell for AI inference is now in limited availability.
Get access