Side-by-side GPU comparison: specs, memory, compute performance, and live cloud pricing.
Verdict
L40 has 24 GB more VRAM, making it better suited for large models and long context windows. For compute-bound workloads like training, L40 delivers 1.5× higher FP16 throughput. At $0.39/hr vs $0.99/hr, L4 is the more cost-efficient choice for inference. L40 supports a broader range of models (624 vs 566 from this catalog), giving more flexibility.
Specifications
| L4 | L40 | |
|---|---|---|
| VRAM | 24 GB | 48 GB |
| VRAM Type | GDDR6 | GDDR6 |
| Memory Bandwidth | 0.3 TB/s | 0.9 TB/s |
| FP16 Performance | 121 TFLOPS | 181 TFLOPS |
| Manufacturer | NVIDIA | NVIDIA |
| FP8 Support | Yes | Yes |
| FP4 Support | No | No |
Price / Performance
Based on cheapest single-GPU on-demand pricing. Lower $/TFLOP = better compute value; lower $/GB = better memory value.
| L4 | L40 | |
|---|---|---|
| $/hr (cheapest) | $0.39✓ best | $0.99 |
| $/TFLOP (compute value) | $0.0032✓ best | $0.0055 |
| $/GB VRAM (memory value) | $0.0163✓ best | $0.0206 |
Cloud Pricing
Cheapest on-demand price per provider (single GPU).
Model Compatibility
Models from the catalog that fit on each GPU, grouped by required precision.
L4 (566 models)
L40 (624 models)
You might also compare…
Pricing data refreshed hourly · Last updated April 11, 2026 · Browse all comparisons