Pricing
Simple, scalable pricing built for AI workloads
On-Demand Pricing
Use it in the way you want. You can access high-performance resources at prices more than 70% cheaper than competitors.
On-Demand Plan
Pay only for what you actually use
Billing calculated on precise instance usage time
Up to 80% savings compared to leading cloud providers
Reserved Plan
Predictable costs with reserved resources
Guaranteed access to reserved computing capacity
Save up to 25% compared to on-demand pricing
Instance | vRAM | CPU | RAM | On-demand | Reserved 1 month | Reserved 3 month |
|---|---|---|---|---|---|---|
G-NHHS-640 8 x H100 SXM | 80GB | 220 vCore | 1920GiB | 64% $65.05/hr $23.18/hr | save 8% $21.45/hr | save 21% $18.27/hr |
G-NHHS-320 4 x H100 SXM | 80GB | 96 vCore | 960GiB | 64% $32.24/hr $11.7/hr | save 9% $10.67/hr | save 23% $9.03/hr |
G-NHHS-80 H100 SXM | 80GB | 24 vCore | 240GiB | 63% $8.06/hr $2.95/hr | save 9% $2.67/hr | save 23% $2.27/hr |
G-NAHP-320 4 x A100 PCIe | 80GB | 64 vCore | 768GiB | 62% $14.49/hr $5.44/hr | save 6% $5.09/hr | save 20% $4.36/hr |
G-NAHP-160 2 x A100 PCIe | 80GB | 32 vCore | 384GiB | 62% $7.25/hr $2.75/hr | save 7% $2.55/hr | save 21% $2.18/hr |
G-NAHP-80 A100 PCIe | 80GB | 16 vCore | 192GiB | 62% $3.62/hr $1.38/hr | save 7% $1.28/hr | save 21% $1.09/hr |
G-NAHP-40 A100 PCIe MIG 3g-40GB | 80GB | 8 vCore | 96GiB | 57% $2.21/hr $0.95/hr | save 18% $0.78/hr | save 25% $0.72/hr |
G-NHHS-640 A100 PCIe MIG 2g-20GB | 80GB | 4 vCore | 48GiB | 62% $1.24/hr $0.48/hr | save 10% $0.43/hr | save 20% $0.38/hr |
G-NHHS-640 A100 PCIe MIG 1g-10GB | 80GB | 2 vCore | 24GiB | 58% $0.55/hr $0.23/hr | save 15% $0.20/hr | save 19% $0.19/hr |
*Storage options are available in 128GB, 256GB, 512GB, 1024GB, and 2048GB increments, with an additional charge of 0.15 won/GB/hour.
ML API Pricing
Choose the right deployment option for your AI workloads with pricing
Serverless
Pay per API call with no infrastructure management
Usage costs calculated based on model type
Dedicated
Dedicated endpoints with exclusive GPU/NPU resources
Billing based on instance running time with flexible scaling options
Serverless Pricing
Text Generation
| Model | Input | Output |
|---|---|---|
HyperCLOVAX-SEED-Text-Instruct-1.5B | $0.015008/1M tokens | $0.025013/1M tokens |
HyperCLOVAX-SEED-Text-Instruct-0.5B | $0.010005/1M tokens | $0.02001/1M tokens |
Trillion-7B-preview | $0.025013/1M tokens | $0.025013/1M tokens |
Speech To Text
| Model | Price |
|---|---|
Whisper Large v3t | $0.00345/60 seconds |
DataHub Pricing
Simple and transparent object storage pricing with no hidden fees or complex tiers.
| Pricing | Unit |
|---|---|
| $0.25 | GiB / Month |
Cost Calculator Example
If you have a variable storage pattern throughout the month, DataHub's hourly calculation ensures you only pay for what you actually use. For example: • 500GiB stored for the first 20 days of the month • 1,500GiB (1.5TiB) stored for the remaining 10 days Your average monthly storage is calculated as: [500GiB × (20 days × 24 hours) + 1,500GiB × (10 days × 24 hours)] ÷ 720 hours = 833GiB-month Monthly cost = 833GiB-month × $0.25 = $208
Ready to accelerate your AI projects?
