GPU MODELS / A10

A10 cloud GPU pricing

21 listings across 11 providers. Cheapest published rate: $0.150/GPU/hr at SpotGPUs. Sorted by per-GPU hourly price. Every row links to the provider's own published price.

WHAT IS THE A10

The A10 fills Ampere's single-slot inference slot: 24 GB, respectable FP16 throughput, 150 W. It sits between the T4/L4 class and the A100 in both capability and price, and remains common in cloud catalogs as a mid-tier serving and graphics option.

MAKER
NVIDIA
ARCHITECTURE
Ampere
LAUNCHED
2021
MEMORY
24 GB GDDR6
MEMORY BANDWIDTH
600 GB/s
TENSOR COMPUTE
125 TFLOPS FP16 tensor dense (250 sparse)
POWER DRAW
150 W
INTERCONNECT
PCIe 4.0 x16

WHAT IT IS GOOD AT

Mid-size inference
7B-13B serving with decent throughput.
Virtual workstations
A common vGPU/VDI card.

WHO SHOULD NOT RENT THE A10

New small-model fleets where the L4 matches it at half the power.
Anything needing more than 24 GB.
ProviderConfigVRAMLocation$/hr$/GPU/hrSource
SpotGPUsSpotGPUsA10 × 124 GBNot published$0.150$0.150receipt
GPUBrazilGPUBrazilA10 × 122 GBNorth America$0.393$0.393receipt
OVHcloudOVHcloudA10 (from) × 124 GBNot published$1.000$1.000receipt
ModalModalA10 × 124 GBNot published$1.102$1.102receipt
CerebriumCerebriumA10 × 124 GBNot published$1.102$1.102receipt
Lambda LabsLambda LabsA10 × 124 GBUS$1.290$1.290receipt
Baidu BCCBaidu BCCA10 × 124 GBChina$1.352$1.352receipt
CTYunCTYunA10 × 124 GBChina$1.377$1.377receipt
Alibaba CloudAlibaba CloudA10 × 824 GBChina (Shanghai)$12.499$1.562receipt
Alibaba CloudAlibaba CloudA10 × 824 GBChina (Hangzhou)$12.499$1.562receipt
Alibaba CloudAlibaba CloudA10 × 824 GBChina (Shenzhen)$12.499$1.562receipt
Alibaba CloudAlibaba CloudA10 × 824 GBChina (Beijing)$12.499$1.562receipt
Alibaba CloudAlibaba CloudA10 × 824 GBUS East (N. Virginia)$12.789$1.599receipt
Kingsoft CloudKingsoft CloudA10 × 124 GBChina$1.801$1.801receipt
Oracle OCIOracle OCIA10 × 124 GBNot published (global list)$2.000$2.000receipt
Alibaba CloudAlibaba CloudA10 × 824 GBSingapore$16.397$2.050receipt
Alibaba CloudAlibaba CloudA10 × 424 GBChina (Shanghai)$8.240$2.060receipt
Alibaba CloudAlibaba CloudA10 × 424 GBChina (Hangzhou)$8.240$2.060receipt
Alibaba CloudAlibaba CloudA10 × 424 GBChina (Beijing)$8.240$2.060receipt
Alibaba CloudAlibaba CloudA10 × 424 GBUS East (N. Virginia)$10.239$2.560receipt
Alibaba CloudAlibaba CloudA10 × 424 GBJapan (Tokyo)$12.258$3.064receipt

THE A10 MARKET, BY THE NUMBERS

PROVIDERS
11
CONFIGURATIONS
21
MEDIAN $/GPU/HR
$1.56
SPREAD
20.4x

11 providers publish 21 A10 configurations on the panel today. Published per-GPU hourly rates run from $0.150 (SpotGPUs) to $3.06 (Alibaba Cloud), a 20x spread between the cheapest and the most expensive published rate for the same chip. The median listing sits at $1.56/GPU/hr. 2 rows were reported in stock by the provider at last observation; the rest are published catalog rates. All rates are on-demand - committed-use and reserved pricing is excluded.

CHEAPEST A10 BY GEOGRAPHY
united statesGPUBrazil$0.393/hrchinaBaidu BCC$1.352/hrsingaporeAlibaba Cloud$2.050/hrjapanAlibaba Cloud$3.064/hr
A10 PRICE TREND, DAILY PANEL SNAPSHOTS

Across every A10 row on the panel, the daily floor moved from $1.000 to $0.150/GPU/hr over 11 days of tracking (-85.0%). The median continuously-tracked listing held flat at $1.29. Snapshots run daily; the full history is public in the repo.

A10 PROVIDER NOTES
SpotGPUs · 1 config · 1 in stock
Sets the panel floor for this model.
from $0.150/hr
GPUBrazil · 1 config · 1 region · 1 in stock
1 of 1 configs reported in stock at last check.
from $0.393/hr
OVHcloud · 1 config
Published catalog rates.
from $1.000/hr
Modal · 1 config
Published catalog rates.
from $1.102/hr
Cerebrium · 1 config
Published catalog rates.
from $1.102/hr
Lambda Labs · 1 config · 1 region
Published catalog rates.
from $1.290/hr
Baidu BCC · 1 config · 1 region
Published catalog rates.
from $1.352/hr
CTYun · 1 config · 1 region
Published catalog rates.
from $1.377/hr
BUYING A10: WHAT THE PANEL SAYS

At the current floor, one A10 running around the clock costs about $110 a month (730 hours of on-demand arithmetic). Node shape matters: the single-GPU floor is $0.150/GPU/hr against $1.562/GPU/hr on an 8-GPU node (+941.6%). Per-GPU floors by config size: 1x $0.15 · 4x $2.06 · 8x $1.56. Compare the alternatives below before committing - the same budget often buys more than one chip class.

A10 VS THE ALTERNATIVES

A10 vs L4
L4 matches the A10's compute at half the power with newer media engines; the A10 counters with 600 GB/s bandwidth. Compare live panel rates.
A10 vs A10G
Same silicon, AWS-specific SKU with vGPU licensing for G5 instances. Performance is equivalent.

Compare live rates: A10 vs T4 · A10 vs L4 · A10 vs A10G · A10 vs A30 · A10 vs A16

A10 PRICING FAQ

What is the cheapest A10 cloud GPU right now?
The cheapest published A10 rate on the panel is $0.150 per GPU-hour at SpotGPUs (A10 x 1, Not published). Every price links to the provider's own published rate.
How much does a A10 cost per hour?
Across 11 providers tracked today, published per-GPU hourly rates for A10 run from $0.15 to $3.06, with a median of $1.56. Committed-use and reserved rates are excluded; everything shown is on-demand.
How many providers rent A10 GPUs?
11 providers publish 21 distinct A10 configurations on Compute Cafe. 2 of those rows were reported in stock by the provider's live endpoint at last observation.
Where is A10 cheapest?
By published per-GPU hourly rate, the cheapest geography for A10 today is united states at $0.39/hr (GPUBrazil). The spread across 4 tracked geographies runs to $3.06/hr in japan.
Are A10 prices going up or down?
Over the 11 days Compute Cafe has tracked the panel, the A10 floor is falling: $1.000 to $0.150/GPU/hr (-85.0%). The median continuously-tracked listing moved from $1.29 to $1.29 (+0.0%). Daily snapshots are public in the repo, so the trend is auditable.
What does a A10 cost per month?
At the current floor of $0.150/GPU/hr, a A10 running 24/7 costs about $110 per month (730 hours of on-demand arithmetic on the cheapest published rate). Committed-use contracts can price lower; the panel tracks on-demand rates only.
Is a full 8-GPU A10 node cheaper than a single card?
On today's panel: the single-GPU floor is $0.150/GPU/hr and the 8-GPU node floor is $1.562/GPU/hr - more expensive per GPU on a full node (+941.6%). Check the table for the exact configs behind each floor.
A10 vs A10G?
The A10G is the AWS-tuned variant of the same chip, built for G5 instances. Treat them as equivalent for performance; compare whichever is cheaper or closer on the panel.

See also alternatives to the A10 · cheapest GPU-hour per model · inference chips · training chips · by location