Cheapest A100 Rental Price — Comparison Table (2026)
A100 80GB SXM4 prices range from $1.89/hr (Vast.ai spot) to $3.20/hr (CoreWeave on-demand). This table compares every major provider. Always verify current pricing — this table is updated weekly.
Provider price comparison
| Provider | Spot $/hr | On-demand $/hr | Notes |
|---|---|---|---|
| Vast.ai | $1.89 | $2.2 | Lowest spot price; variable availability |
| RunPod | $2.09 | $2.49 | Reliable; community + secure cloud tiers |
| Lambda Labs | — | $2.5 | On-demand only; good reliability |
| Hyperstack | — | $2.2 | European regions available; NVLink clusters |
| CoreWeave | — | $3.2 | Enterprise SLAs; highest reliability |
Verified GPU rental prices across all major providers. Start in minutes, pay by the hour.
Frequently Asked Questions
What is the cheapest way to rent an A100?
Vast.ai spot instances offer the lowest A100 prices (~$1.89/hr) but availability fluctuates. For reliable access, RunPod community cloud ($2.09/hr spot) is a good balance. For production workloads needing SLAs, CoreWeave or Lambda are more appropriate despite higher prices.
A100 40GB vs 80GB — when does the extra VRAM matter?
An A100 80GB is worth the premium (~30% more) when running: (1) models larger than 20B at FP16, (2) large context windows with big KV-cache, (3) large batch inference for throughput. For models that fit in 40GB (< 20B at FP16), the 40GB is more cost-efficient.
Data verified 2026-08-20. GPU rental prices change frequently — verify before purchasing. See methodology.