Accelerators pricing
AWS Inferentia prices in ap-south-1 (Asia Pacific (Mumbai))
Updated 2026-10-19
inf1.xlarge $0.24/h, inf1.2xlarge $0.381/h, inf1.6xlarge $1.24/h, inf1.24xlarge $4.96/h, inf2.xlarge $0.9857/h, inf2.8xlarge $2.56/h, inf2.48xlarge $16.88/h.
Sizes
| Instance | Accelerators | Accel. memory | On-demand $/h | $/chip-hr | 3-yr RI no upfront $/h | Hours per $25k |
|---|---|---|---|---|---|---|
| inf1.xlarge | Inferentia ×1 | n/a | $0.24 | $0.24 | $0.115 | 104,167 h |
| inf1.2xlarge | Inferentia ×1 | n/a | $0.381 | $0.381 | $0.183 | 65,617 h |
| inf1.6xlarge | Inferentia ×4 | n/a | $1.24 | $0.3103 | $0.597 | 20,145 h |
| inf1.24xlarge | Inferentia ×16 | n/a | $4.96 | $0.3103 | $2.39 | 5,035 h |
| inf2.xlarge | Inferentia2 ×1 | 32 GB | $0.9857 | $0.9857 | $0.4586 | 25,364 h |
| inf2.8xlarge | Inferentia2 ×1 | 32 GB | $2.56 | $2.56 | $1.19 | 9,772 h |
| inf2.48xlarge | Inferentia2 ×12 | 384 GB | $16.88 | $1.41 | $7.85 | 1,481 h |
GPU instances in ap-south-1 for comparison
| Instance | Accelerators | Accel. memory | On-demand $/h | $/chip-hr | 3-yr RI no upfront $/h | Hours per $25k |
|---|---|---|---|---|---|---|
| g4dn.xlarge | NVIDIA T4 Tensor Core ×1 | 16 GB | $0.579 | $0.579 | $0.281 | 43,178 h |
| g5.xlarge | NVIDIA A10G ×1 | 24 GB | $1.21 | $1.21 | $0.5219 | 20,695 h |
| g6.xlarge | NVIDIA L4 ×1 | 24 GB | $0.9664 | $0.9664 | $0.4436 | 25,869 h |
| p4d.24xlarge | NVIDIA A100 ×8 | 320 GB | $26.34 | $3.29 | $14.18 | 949 h |
| p5.48xlarge | NVIDIA H100 ×8 | 640 GB | $66.05 | $8.26 | $28.53 | 379 h |
Other regions
Related cost calculators
- EC2 Pricing Calculator · Compare Instance Types & Costs
- Amazon SageMaker AI Pricing Calculator · On-Demand Studio, Training & Inference
Prices and credit amounts are estimates and may change. Check AWS for current terms.