Loading page
performance tier
Mid-Tier GPUs Cloud Pricing
Mid-tier GPUs balance price and capability for production inference, moderate training, and professional workloads. This tier includes popular consumer cards (RTX 3080, RTX 4070) and datacenter options (A10G, A30). VRAM ranges from 8–64 GB. They're commonly used for serving ML models in production and fine-tuning with parameter-efficient methods.
GPUs 25
Providers 18
From $0.030/hr
Mid-Tier GPUs Available in the Cloud
A16
Server64 GBAmpereNVIDIA
$0.510/hr3 providers
Quadro RTX 5000
16 GBTuringNVIDIA
$0.820/hr1 provider
Quadro RTX 6000
24 GBTuringNVIDIA
$0.130/hr2 providers
RTX 2080 Ti
11 GBTuringNVIDIA
$0.050/hr3 providers
RTX 3060 Ti
8 GBAmpereNVIDIA
$0.030/hr4 providers
RTX 3070
8 GBAmpereNVIDIA
$0.040/hr5 providers
RTX 3070 Ti
8 GBAmpereNVIDIA
$0.060/hr3 providers
RTX 3080
10 GBAmpereNVIDIA
$0.050/hr6 providers
RTX 3090
24 GBAmpereNVIDIA
$0.080/hr7 providers
RTX 4000 Ada
20 GBAda LovelaceNVIDIA
$0.160/hr7 providers
RTX 4060 Ti
8 GBAda LovelaceNVIDIA
$0.076/hr3 providers
RTX 4070
12 GBAda LovelaceNVIDIA
$0.070/hr4 providers
RTX 4070 SUPER
12 GBAda LovelaceNVIDIA
No recent pricing
RTX 4080
16 GBAda LovelaceNVIDIA
$0.110/hr5 providers
RTX 4500 Ada
24 GBAda LovelaceNVIDIA
$0.270/hr1 provider
RTX 5060 Ti
16 GBBlackwellNVIDIA
$0.060/hr5 providers
RTX 5070
12 GBBlackwellNVIDIA
$0.070/hr5 providers
RTX A4500
Server20 GBAmpereNVIDIA
$0.088/hr3 providers
RTX A5000
24 GBAmpereNVIDIA
$0.070/hr8 providers
RX 7800 XT
16 GBRDNA 3AMD
$0.042/hr1 provider
RX 7900 XT
20 GBRDNA 3AMD
$0.054/hr1 provider
RX 7900 XTX
24 GBRDNA 3AMD
$0.066/hr1 provider
RX 9060 XT
16 GBRDNA 4AMD
$0.049/hr1 provider
RX 9070 XT
16 GBRDNA 4AMD
$0.070/hr1 provider
Titan RTX
24 GBTuringNVIDIA
$0.120/hr2 providers
Sample Mid-Tier GPUs Pricing
| Provider | Price / hr |
|---|---|
$0.050/hr 1× | |
$0.160/hr 1× | |
$0.165/hr 2× | |
$0.250/hr 1× | |
$0.440/hr 8× | |
$0.561/hr 2× | |
$0.760/hr 1× | |
$1.56/hr 1× | |
$0.160/hr 1× |
Direct from providerVia marketplace
Showing 9 of 120 price points. Visit individual GPU pages above for full pricing.
Frequently Asked Questions
What workloads suit mid-tier GPUs?
Mid-tier GPUs handle production inference for models up to 13B parameters, fine-tuning with LoRA/QLoRA, batch processing, image generation, and video encoding. They offer a practical balance between cost and throughput.
How do mid-tier GPUs compare to high-tier for inference?
Mid-tier GPUs have lower memory bandwidth and fewer tensor cores, so throughput per GPU is lower. However, they can be more cost-effective for workloads that don't need the full power of high-tier cards. Compare pricing per token of output above.