Low-power inference accelerator for cost-sensitive cloud deployments.
Weekly median price per GPU per hour · Get the data
By provider shows one card per company, placed where its best offer ranked. By configuration lists every offer.
What you can rent comes first: in stock, then waitlist, then not reported, then out of stock. Priced offers rank ahead of quote-only, and on-demand ahead of other billing types.
Within a group, five factors set the order:
We match the provider, its country, GPU model, form factor, billing type, availability and instance name. Matching is partial and case-insensitive. Everyday words work too, so "interruptible" finds spot and "sold out" finds out of stock.
No providers match .
Heads up: A provider's own page may quote a different figure, for example on tax, region or a promotion. A monthly-only plan shows a derived hourly rate. Verify before provisioning. More on how we price.
One L4 has 24 GB of VRAM. In practice, that's enough memory for roughly 28B parameters at 4-bit or 6B at 16-bit, assuming a 32K context. Below are some open-weight LLMs, with the estimated memory and GPUs each one needs.
| Model | Memory (INT4 / FP4) | L4s needed | Cost /hr | Cost /mo |
|---|---|---|---|---|
Qwen3.8-27B
27B
|
18 GB
|
1
|
$0.87
|
$626
|
Gemma 4 31B
30.7B
|
22 GB
|
1
|
$0.87
|
$626
|
GPT-OSS-120B
117B · 5.1B active
|
67 GB
|
4
|
$3.48
|
$2,506
|
GLM-5.3-Flash
320B · 18B active
|
178 GB
|
9
|
–
|
–
|
MiniMax-M3
428B · 23B active
|
239 GB
|
12
|
–
|
–
|
DeepSeek V4.1 Flash
552B · 16B active
|
306 GB
|
15
|
–
|
–
|
Kimi K3
2.8T · 104B active
|
1,544 GB
|
72
|
–
|
–
|
Estimates based on the median on-demand rate. Memory is weights plus FP8 KV cache at 32K context per request (FP16/BF16 in the 16-bit column). GPU counts assume 90% of advertised VRAM is usable. No guarantee of runtime support, usable performance, or that a matching quantized build exists. Pricing methodology.
Nvidia L4 · Per GPU
| Compute · dense | |
|---|---|
| FP8 | 242 TFLOPS 485 with sparsity |
| FP16 / BF16 | 121 TFLOPS 242 with sparsity |
| INT8 | 242 TOPS 485 with sparsity |
| FP32 | 30.3 TFLOPS |
| Precision support | FP8FP16BF16TF32FP32INT4INT8 |
| Memory | |
|---|---|
| Capacity | 24 GB GDDR6 |
| Bandwidth | 300 GB/s |
| Bus width | 192-bit |
| ECC | Yes |
| Silicon | |
|---|---|
| Architecture | Ada Lovelace |
| Process | TSMC 4N |
| Shader cores | 7,424 CUDA cores |
| Matrix cores | 232 Tensor cores |
| Compute units | 58 SMs |
| Fabric and host | |
|---|---|
| Host interface | PCIe 4.0 x16 |
| Power | |
|---|---|
| Board power | 72 W |
| Cooling | Passive |
Source: official Nvidia L4 datasheet.
As of October 18, 2026, the median on-demand price is $0.87 per GPU per hour across 17 providers with a priced on-demand config.
| Billing type | Configs | Median /GPU/hr | Cheapest |
|---|---|---|---|
On-demand | 73 | $0.87 | $0.33 (in stock, Theta EdgeCloud) |
Reserved | 83 | $0.52 | $0.32 (36 mo, not verified in stock, Google Cloud) |
Spot | 23 | $0.29 (in stock, Jarvislabs) |
No median for spot: we only show this when at least 3 providers list the GPU on that billing type.
The cheapest verified in-stock on-demand L4 tracked by GetDeploying is $0.33 per GPU per hour from Theta EdgeCloud. Other billing types: reserved $0.32 (36 mo, not verified in stock) from Google Cloud and spot $0.29 (in stock) from Jarvislabs.
At 720 hours per month, one L4 costs an estimated $626 at the median on-demand price. Cheapest verified in stock: $238 per month on-demand, $288 reserved, $209 spot.
GetDeploying currently tracks L4 configs from 21 providers. The cheapest verified in-stock on-demand configs come from Theta EdgeCloud, Jarvislabs, GPU.ai and Runpod. See the full price comparison above for every provider and config.
As of October 18, 2026, the median on-demand price has been flat over the past 90 days, at about $0.87 per GPU per hour, though it is about 6% above where it was a year ago.
One L4 runs models up to roughly 28B parameters at 4-bit quantization or 6B at 16-bit, assuming a 32K context. Larger models run across multiple GPUs: the model table above shows the estimated memory and GPU count for popular open-weight LLMs.
24GB GDDR6 in a low-power 72W form factor. FP8 support for efficient inference. Excellent performance-per-watt for deploying AI models at scale in cost-sensitive environments.
Limited VRAM caps model sizes at ~13B parameters (FP16). Built for inference, not training. Low memory bandwidth (300 GB/s) may limit throughput on memory-bound workloads.
Last updated
