COMPARE / A10G VS L4

A10G vs L4

Specs from the chip makers' datasheets; prices from the live Compute Cafe panel, refreshed hourly. Every rate is an on-demand published price with a source receipt.

CHEAPEST A10G
$1.000/hr
Hugging Face · median $1.25 · 4 providers
CHEAPEST L4
$0.080/hr
SpotGPUs · median $0.70 · 26 providers
CHEAPER RIGHT NOW
L4
live panel, 2026-09-27
A10GL4
MAKERNVIDIANVIDIA
ARCHITECTUREAmpereAda Lovelace
LAUNCHED20212023
MEMORY24 GB GDDR624 GB GDDR6
MEMORY BANDWIDTH600 GB/s300 GB/s
TENSOR COMPUTE125 TFLOPS FP16 tensor dense (250 sparse)121 TFLOPS FP16 tensor dense (242 sparse); 242 TOPS INT8 dense
POWER DRAW150 W72 W
INTERCONNECTPCIe 4.0 x16PCIe 4.0 x16

HOW TO CHOOSE

A10G vs L4
AWS's own newer inference chip (G6 instances): similar compute, newer engines, often better $/hr. Check the panel's live AWS rows for both.
Workload fit
A10G: mid-size inference, graphics workstations. L4: small-model serving, video pipelines, edge and burst inference.

A10G VS L4 - FAQ

Which is cheaper to rent, A10G or L4?
By today's published on-demand rates: A10G starts at $1.000/GPU/hr (median $1.25, 4 providers) and L4 starts at $0.080/GPU/hr (median $0.70, 26 providers). The cheapest live listing is the L4 at SpotGPUs. Figures recompute from the panel on every refresh.
A10G or L4 - which should I rent?
A10G: The A10G is the AWS-specific variant of the A10, deployed in G5 instances with vGPU support. L4: The L4 is the modern successor to the T4: a 72 W single-slot card with roughly double its predecessor's throughput, FP8 support, and Ada's AV1 video engines. The side-by-side spec table and live price rows above give the full picture.
Where can I rent A10G and L4?
4 providers publish A10G rates and 26 publish L4 rates on Compute Cafe. Every row links to the provider's own published price - committed-use and reserved rates are excluded.

Full pricing: A10G · L4 · all comparisons