Two Nvidia data center GPUs, one Ada Lovelace and one Volta: 24 GB GDDR6 against 32 GB HBM2, 3x the memory bandwidth on the V100, and NVLink 300 GB/s on the V100 where the L4 is PCIe only.
At a glance
Updated 18 minutes ago- Memory
- 24 GB
- GDDR6
- Memory bandwidth
- 300 GB/s
- From
- $0.31 /GPU/hr
- 19 providers
- Memory
- 32 GB
- HBM2
- Memory bandwidth
- 900 GB/s
- From
- $0.11 /GPU/hr
- 20 providers
Price comparison
Cheapest listed price per GPU per hour by billing type, and the median across providers.
| Billing | L4 | V100 | Difference |
|---|---|---|---|
| On-demand | $0.31 | $0.11 | 2.9x more expensive on L4 |
| Spot | $0.13 | $0.06 | 2.4x more expensive on L4 |
| Reserved | $0.32 | $0.12 | 2.5x more expensive on L4 |
| Median | $0.90 | $0.99 | 10% cheaper on L4 |
Median: Median across providers, one vote each, so a large catalog does not outweigh a small one.
Price history
L4 vs V100 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | L4 | V100 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 24 GB GDDR6 | 32 GB HBM2 | 1.3x more on V100 |
| Bandwidth | 300 GB/s | 900 GB/s | 3x higher on V100 |
| Bus width | 192-bit | 4,096-bit | 21x wider on V100 |
| Compute · dense | |||
| FP8 | 242 TFLOPS | Not supported | L4 only |
| FP16 / BF16 | 121 TFLOPS | 125 TFLOPS | Within 5% |
| INT8 | 242 TOPS | ||
| FP32 | 30.3 TFLOPS | 15.7 TFLOPS | 1.9x higher on L4 |
| FP64 | Not supported | 7.8 TFLOPS | V100 only |
| Platform | |||
| Interconnect | PCIe only | NVLink 300 GB/s | NVLink on V100 only |
| Board power | 72 W | 300 W | 4.2x higher on V100 |
| Released | Q1 2023 | Q2 2017 | 6 years newer on L4 |
| Precision | |||
| Supported formats | FP8FP16BF16TF32FP32INT4INT8 | FP16FP32FP64INT8 | |
Figures are for each model's headline configuration. Where a form factor carries a different number, it is named under the figure.
Compute: Vendor peak figures, dense and per GPU. Not measured throughput.
Source: L4 datasheet and V100 datasheet.
Where to rent them
Providers listing these GPUs, those with both first, so you can switch without moving clouds.
| Provider | L4 /GPU/hr | V100 /GPU/hr |
|---|---|---|
Vast.ai
|
On-Demand $0.31 | On-Demand $0.11 |
Runcrate
|
On-Demand $1.05 | On-Demand $2.57 |
Sesterce
|
On-Demand $1.04 | On-Demand $2.57 |
GPU.ai
|
On-Demand $0.32 | -- |
Runpod
|
On-Demand $0.49 | -- |
UpCloud
|
On-Demand $0.69 | -- |
Scaleway
|
On-Demand $0.91 | -- |
Hexabyte
|
On-Demand $0.42 | -- |
HyperAI
|
-- | On-Demand $0.14 |
Verda
|
-- | On-Demand $0.17 |
Geodd
|
-- | On-Demand $0.16 |
TensorDock
|
-- | On-Demand $0.28 |
Cyfuture AI
|
-- | On-Demand $0.55 |
19 providers list the L4 and 20 the V100. Every listing is on the L4 page and the V100 page.
Similar GPUs
Other comparisons in the same class.
Our data for Nvidia L4 was last updated on Sept. 7, 2026, and for Nvidia V100 on Sept. 7, 2026.