NVIDIA L4
InferenceNvidia · Ada Lovelace · 2023
Single-slot, low-profile PCIe card; 72W TDP, 2x NVENC / 4x NVDEC. Designed as an efficient inference/video accelerator. fp16Tflops/fp8Tflops are NVIDIA's published Tensor Core peak figures "with sparsity" (dense throughput is approximately half).
- VRAM
- 24 GB
- Memory
- GDDR6
- FP16
- 242 TF
- FP8
- 485 TF
From
Verified 8 Jul 2026€0.38/GPU·h
Offerings of this GPU
4 rows
| Provider | Config | Regions | €/GPU·h | Availability | Sov. | Verified | |
|---|---|---|---|---|---|---|---|
| Seeweb Compute | 1× GPU · VM Cloud Server GPU - L4 | 🇮🇹 IT | €0.38/GPU·h | Available | 63 | 2026-07-08 | |
| Scaleway Compute | 8× GPU · VM L4-8-24G | 🇫🇷 FR | €0.79/GPU·h egress free | Available | 69 | 2026-07-08 | |
| OVHcloud Compute | 1× GPU · VM Public Cloud GPU L4 | 🇫🇷 FR | Quote only egress free | On request | 80 | 2026-07-08 | |
| Sesterce Compute | 1× GPU · Bare metal L4 | 🇫🇷 FR | Quote only | Available | 60 | 2026-07-08 |
Spec sources
- Official Docsresources.nvidia.com/en-us-data-center-overview/l4-gpu-datasheet8 Jul 2026
- Official Docswww.nvidia.com/en-us/data-center/l48 Jul 2026