GPU MODELS / RTX 4080

RTX 4080 cloud GPU pricing

15 listings across 8 providers. Cheapest published rate: $0.110/GPU/hr at Salad. Sorted by per-GPU hourly price. Every row links to the provider's own published price.

WHAT IS THE RTX 4080

The RTX 4080 is Ada's second-tier consumer flagship: fast, efficient, and capped at 16 GB. It serves the same dev-and-small-model niche as the newer 5080 at last-generation prices - often the cheaper rental for identical workloads.

MAKER
NVIDIA
ARCHITECTURE
Ada Lovelace
LAUNCHED
2022
MEMORY
16 GB GDDR6X
MEMORY BANDWIDTH
717 GB/s
TENSOR COMPUTE
48.7 TFLOPS FP32; 4th-gen tensor cores
POWER DRAW
320 W
INTERCONNECT
PCIe 4.0 x16

WHAT IT IS GOOD AT

Small-model inference
Strong - 7B-13B quantized models serve well on 717 GB/s.
Media generation
Good fit for diffusion pipelines at moderate batch sizes.
Fine-tuning
LoRA to ~13B; the 16 GB ceiling rules out fuller runs.

WHO SHOULD NOT RENT THE RTX 4080

Models past ~13B - 16 GB is the wall; the 4090's 24 GB is the consumer step up.
FP4/FP8-native stacks, which want Blackwell.
ProviderConfigVRAMLocation$/hr$/GPU/hrSource
SpotGPUsSpotGPUsRTX 4080 × 116 GBNot published$0.090$0.090STALEreceipt
SaladSaladRTX 4080 × 116 GBGlobal (distributed)$0.110$0.110receipt
QuickPodQuickPodRTX 4080 × 116 GBNot published$0.130$0.130STALEreceipt
CloreCloreRTX 4080 × 116 GBIT$0.187$0.187receipt
RunPodRunPodRTX 4080 × 116 GBGlobal$0.270$0.270receipt
gpu.aigpu.aiRTX 4080 × 116 GBNot published$0.290$0.290receipt
Hostnot GPUHostnot GPURTX 4080 × 116 GBAsia-Pacific$0.320$0.320receipt
Hostnot GPUHostnot GPURTX 4080 × 216 GBUS West$0.680$0.340receipt
Hostnot GPUHostnot GPURTX 4080 × 116 GBUS West$0.350$0.350receipt
Hostnot GPUHostnot GPURTX 4080 × 216 GBUS East$0.710$0.355receipt
Hostnot GPUHostnot GPURTX 4080 × 116 GBUS East$0.360$0.360receipt
Hostnot GPUHostnot GPURTX 4080 × 816 GBCanada Central$3.760$0.470receipt
Hostnot GPUHostnot GPURTX 4080 × 416 GBCanada Central$1.890$0.472receipt
GPUBrazilGPUBrazilRTX 4080 × 116 GBNorth America$0.531$0.531receipt
GPUBrazilGPUBrazilRTX 4080 × 116 GBAsia-Pacific$0.531$0.531receipt

THE RTX 4080 MARKET, BY THE NUMBERS

PROVIDERS
8
CONFIGURATIONS
15
MEDIAN $/GPU/HR
$0.35
SPREAD
4.8x

8 providers publish 15 RTX 4080 configurations on the panel today. Published per-GPU hourly rates run from $0.110 (Salad) to $0.53 (GPUBrazil), a 5x spread between the cheapest and the most expensive published rate for the same chip. The median listing sits at $0.35/GPU/hr. 10 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 RTX 4080 BY GEOGRAPHY
italyClore$0.187/hrunited statesHostnot GPU$0.340/hrcanadaHostnot GPU$0.470/hr
RTX 4080 PRICE TREND, DAILY PANEL SNAPSHOTS

Across every RTX 4080 row on the panel, the daily floor held flat at $0.110/GPU/hr over 53 days of tracking. Snapshots run daily; the full history is public in the repo.

RTX 4080 PROVIDER NOTES
Salad · 1 config · 1 region
Sets the panel floor for this model.
from $0.110/hr
Clore · 1 config · 1 region · 1 in stock
1 of 1 configs reported in stock at last check.
from $0.187/hr
RunPod · 1 config · 1 region
Published catalog rates.
from $0.270/hr
gpu.ai · 1 config
Published catalog rates.
from $0.290/hr
Hostnot GPU · 7 configs · 4 regions · 7 in stock
Widest regional footprint here - 4 regions.
from $0.320/hr
GPUBrazil · 2 configs · 2 regions · 2 in stock
2 of 2 configs reported in stock at last check.
from $0.531/hr
BUYING RTX 4080: WHAT THE PANEL SAYS

At the current floor, one RTX 4080 running around the clock costs about $80 a month (730 hours of on-demand arithmetic). Node shape matters: the single-GPU floor is $0.110/GPU/hr against $0.470/GPU/hr on an 8-GPU node (+327.3%). Per-GPU floors by config size: 1x $0.11 · 2x $0.34 · 4x $0.47 · 8x $0.47. Compare the alternatives below before committing - the same budget often buys more than one chip class.

RTX 4080 VS THE ALTERNATIVES

RTX 4080 vs RTX 5080
Same 16 GB, newer architecture, roughly 15% faster. For most rental workloads they are interchangeable - take whichever the panel prices lower.
RTX 4080 vs RTX 4090
The 4090's 24 GB fits whole model classes the 4080 cannot. For LLM work the premium is usually worth it.

Compare live rates: RTX 4080 vs RTX 5080 · RTX 4080 vs RTX 4090 · RTX 4080 vs RTX 4080 Super

RTX 4080 PRICING FAQ

What is the cheapest RTX 4080 cloud GPU right now?
The cheapest published RTX 4080 rate on the panel is $0.110 per GPU-hour at Salad (RTX 4080 x 1, Global (distributed)). Every price links to the provider's own published rate.
How much does a RTX 4080 cost per hour?
Across 8 providers tracked today, published per-GPU hourly rates for RTX 4080 run from $0.11 to $0.53, with a median of $0.35. Committed-use and reserved rates are excluded; everything shown is on-demand.
How many providers rent RTX 4080 GPUs?
8 providers publish 15 distinct RTX 4080 configurations on Compute Cafe. 10 of those rows were reported in stock by the provider's live endpoint at last observation.
Where is RTX 4080 cheapest?
By published per-GPU hourly rate, the cheapest geography for RTX 4080 today is italy at $0.19/hr (Clore). The spread across 3 tracked geographies runs to $0.47/hr in canada.
Are RTX 4080 prices going up or down?
Over the 53 days Compute Cafe has tracked the panel, the RTX 4080 floor is holding flat: $0.110 to $0.110/GPU/hr (+0.0%). Daily snapshots are public in the repo, so the trend is auditable.
What does a RTX 4080 cost per month?
At the current floor of $0.110/GPU/hr, a RTX 4080 running 24/7 costs about $80 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 RTX 4080 node cheaper than a single card?
On today's panel: the single-GPU floor is $0.110/GPU/hr and the 8-GPU node floor is $0.470/GPU/hr - more expensive per GPU on a full node (+327.3%). Check the table for the exact configs behind each floor.
Is the RTX 4080 still worth renting?
Yes when it prices below the 5080 - same memory, nearly the same real-world speed for small-model serving and dev work.

See also alternatives to the RTX 4080 · cheapest GPU-hour per model · inference chips · training chips · by location