Budget inference GPUs

Lower-cost GPU listings with at least 16 GB of VRAM, priced for inference workloads. Sorted by price.

ProviderGPUCOUNTVRAM TOTAL$/HRSPOT$/MOCOMPLIANCESTATUSAction
VastCHEAPESTRTX 3090BG
1x
Show all 23 configurations
×1Latvia, LV$0.16/hr$0.16/hrLaunch ↗
×1Thailand, TH$0.16/hr$0.16/hrLaunch ↗
×1 · 13 GB RAMBG$0.16/hr$0.16/hrLaunch ↗
×1Kosovo, XK$0.16/hr$0.16/hrLaunch ↗
×1United Arab Emirates, AE$0.16/hr$0.16/hrLaunch ↗
×1 · 63 GB RAMUS$0.16/hr$0.16/hrLaunch ↗
×1Türkiye, TR$0.16/hr$0.16/hrLaunch ↗
×1Norway, NO$0.16/hr$0.16/hrLaunch ↗
×1Washington, US$0.16/hr$0.16/hrLaunch ↗
×1Tennessee, US$0.16/hr$0.16/hrLaunch ↗
×1Utah, US$0.17/hr$0.17/hrLaunch ↗
×1Vietnam, VN$0.17/hr$0.17/hrLaunch ↗
×1Samarskaya oblast', RU$0.18/hr$0.18/hrLaunch ↗
×1Texas, US$0.18/hr$0.18/hrLaunch ↗
×1Wyoming, US$0.18/hr$0.18/hrLaunch ↗
×1Ukraine, UA$0.18/hr$0.18/hrLaunch ↗
×1British Columbia, CA$0.18/hr$0.18/hrLaunch ↗
×1Ukraine, UA$0.18/hr$0.18/hrLaunch ↗
×1Japan, JP$0.18/hr$0.18/hrLaunch ↗
×1Sverdlovskaya oblast', RU$0.18/hr$0.18/hrLaunch ↗
×1Moskovskaya oblast', RU, Moskva, RU, Virginia, US$0.18/hr$0.18/hrLaunch ↗
×1Oman, OM$0.18/hr$0.18/hrLaunch ↗
×1California, US$0.18/hr$0.18/hrLaunch ↗
24GB
$0.16/hr
$0.11/hr$116.80/moUnavailRent
Ai GalaxyCHEAPESTV100CN
1x
16GB
$0.16/hr
$116.80/moUnavailRent
VastCHEAPESTV100Bulgaria, BG
1x
Show all 4 configurations
×1Colorado, US$0.16/hr$0.16/hrLaunch ↗
×1Sri Lanka, LK$0.16/hr$0.16/hrLaunch ↗
×1Virginia, US$0.18/hr$0.18/hrLaunch ↗
×1Bulgaria, BG$0.18/hr$0.18/hrLaunch ↗
16GB
$0.16/hr
$0.09/hr$116.80/moUnavailRent
VastCHEAPESTRTX A4000Iceland, IS
1x
16GB
$0.16/hr
$0.15/hr$116.80/moUnavailRent
RunPodCHEAPESTRTX A5000Region n/a
1x
24GB
$0.16/hr
$116.80/moUnavailRent
RunPodCHEAPESTNVIDIA RTX A5000Region n/a
1x
24GB
$0.16/hr
$116.80/moUnavailRent
Alibaba CloudT4ap-northeast-1
1x
Show all 8 configurations
×1 · 15 GB RAMus-east-1$0.17/hr$0.17/hrLaunch ↗
×1 · 15 GB RAMus-west-1$0.17/hr$0.17/hrLaunch ↗
×1 · 15 GB RAMcn-hongkong$0.17/hr$0.17/hrLaunch ↗
×1 · 15 GB RAMeu-central-1$0.18/hr$0.18/hrLaunch ↗
×1 · 15 GB RAMap-southeast-5$0.18/hr$0.18/hrLaunch ↗
×1 · 15 GB RAMap-northeast-1$0.18/hr$0.18/hrLaunch ↗
×1 · 31 GB RAMus-east-1$0.19/hr$0.19/hrLaunch ↗
×1 · 15 GB RAMap-southeast-3$0.19/hr$0.19/hrLaunch ↗
16GB
$0.17/hr
$124.10/moUnavailRent
RunPodRTX A4000Global
1x
Show all 3 configurations
×1n/a$0.17/hr$0.17/hrLaunch ↗
×1Global$0.17/hr$0.17/hrLaunch ↗
×1n/a$0.17/hr$0.17/hrLaunch ↗
16GB
$0.17/hr
$124.10/moUnavailRent
VastL4The Netherlands, NL
1x
24GB
$0.17/hr
$0.13/hr$124.10/moUnavailRent
LiumRTX 3090Germany
1x
24GB
$0.18/hr
$131.40/moUnavailRent
AutoDLRTX 3090CN
1x
24GB
$0.18/hr
$131.40/moUnavailRent
RunPodNVIDIA RTX 4000 SFF Ada GenerationRegion n/a
1x
20GB
$0.18/hr
$131.40/moUnavailRent
RunPodRTX 4000 Ada SFFRegion n/a
1x
20GB
$0.18/hr
$131.40/moUnavailRent
VastRTX 5060 TiDE
1x
16GB
$0.18/hr
$131.40/moUnavailRent
RunPodV100Region n/a
1x
16GB
$0.19/hr
$138.70/moUnavailRent
Alibaba CloudA10cn-heyuan
1x
24GB
$0.19/hr
$138.70/moUnavailRent

Spot = gpufinder-published price strictly below on-demand — gpufinder is the only feed that differentiates spot. '—' means the provider doesn't publish a differential spot price (unknown, not zero).