A10G vs A2
Specifications, today's prices from 6 providers, price per unit of memory and compute, and who rents each cheapest. Better values are highlighted; lower is better for prices.
Specs, prices, value
Highlighted cells are the better of the two: more memory and compute, lower prices, lower cost per unit.
| A10G | A2 | Difference | |
|---|---|---|---|
| Memory & compute | |||
| VRAM | 24 GB | 16 GB | +50% |
| Memory bandwidth | — | 0.20 TB/s | — |
| FP16 tensor | — | 18 TFLOPS | — |
| FP32 | — | 4.5 TFLOPS | — |
| Power | — | 60 W | — |
| Prices today | |||
| Cheapest on-demand | $1.00 | $0.16 | +512% |
| Median on-demand | $1.01 | $0.23 | +341% |
| Highest on-demand | $2.45 | $0.43 | +471% |
| Cheapest spot | $0.24 | $0.13 | +87% |
| Cheapest reserved | — | $0.18 | — |
| Current offers (n) | 11 | 3 | +267% |
| Providers with a fresh offer | 3 | 3 | 0% |
| Value | |||
| Per GB of VRAM · hour (cheapest) | $0.042 | $0.010 | +308% |
| Per GB of VRAM · hour (median) | $0.042 | $0.014 | +194% |
| Per PFLOP FP16 · hour (median) | — | $12.661 | — |
| Per TB/s · hour (median) | — | $1.140 | — |
| Trend | |||
| 30-day median change | — | — | — |
| 90-day median change | — | — | — |
Cheapest providers today · on-demand · current offersWho rents each one cheapest
- 1
Hugging Face1× $1.00 - 2
AWSTrust 791× $1.01 - 3
BasetenTrust 871× $1.21
- 1
Infomaniak1× $0.16 - 2
Cherry ServersTrust 821× $0.23
- 3
Selectel1× $0.43
Choose one, choose the other
Choose the A10G if…
- your model needs more than 16 GB per GPU: 24 GB against 16 GB.
Choose the A2 if…
- the hourly price matters most: median $0.23 against $1.01, 77% less.
Questions about A10G vs A2
Which is cheaper to rent, the A10G or the A2?
Today the cheapest current on-demand A10G is $1.00 per GPU-hour and the cheapest A2 is $0.16; medians are $1.01 and $0.23. Cheapest excludes offers marked unavailable and counts one floor per marketplace.
How much memory do the A10G and A2 have?
24 GB of memory on the A10G, 16 GB of GDDR6 on the A2.
Where do these numbers come from?
From 14 current offers read on the providers' own pages and APIs over the last 45 days (reference 00:32 UTC), or in our October 2026 market study for the few providers our collector does not read. Specifications come from manufacturer datasheets. Every price on the GPU pages links to its receipt and shows its read date.
