Comparison · GPU vs GPU · read 00:32 UTCL4 vs T4
Specifications, today's prices from 51 providers, price per unit of memory and compute, and who rents each cheapest. Better values are highlighted; lower is better for prices.
Side by sideSpecs, prices, value
Highlighted cells are the better of the two: more memory and compute, lower prices, lower cost per unit.
| L4 | T4 | Difference | |
|---|---|---|---|
| Memory & compute | |||
| VRAM | 24 GB | 16 GB | +50% |
| Memory bandwidth | 0.30 TB/s | 0.32 TB/s | -6% |
| FP16 tensor | 121 TFLOPS | 65 TFLOPS | +86% |
| FP32 | 30.3 TFLOPS | 8.1 TFLOPS | +274% |
| Power | 72 W | 70 W | +3% |
| Prices today | |||
| Cheapest on-demand | $0.27 | $0.20 | +39% |
| Median on-demand | $0.80 | $0.49 | +63% |
| Highest on-demand | $2.01 | $1.22 | +65% |
| Cheapest spot | $0.10 | $0.07 | +45% |
| Cheapest reserved | $0.32 | $0.17 | +83% |
| Current offers (n) | 72 | 87 | -17% |
| Providers with a fresh offer | 25 | 26 | -4% |
| Value | |||
| Per GB of VRAM · hour (cheapest) | $0.011 | $0.012 | -7% |
| Per GB of VRAM · hour (median) | $0.034 | $0.031 | +9% |
| Per PFLOP FP16 · hour (median) | $6.651 | $7.580 | -12% |
| Per TB/s · hour (median) | $2.683 | $1.540 | +74% |
| Trend | |||
| 30-day median change | -1.5% | -13.3% | — |
| 90-day median change | -2.8% | -13.3% | — |
90 days · median on-demand price · rebased to 100How the two have moved
Both medians indexed to 100 ninety days ago, so the slopes compare even though the price levels differ.
Cheapest providers today · on-demand · current offersWho rents each one cheapest
- 1
Infomaniak2× $0.27 - 2
Hikube1× $0.27 - 3
OpenRelay1× $0.43
- 1
iwinv1× $0.20 - 2
Theta EdgeCloud1× $0.20 - 3
LiumTrust 401× $0.22
Verdict · computed from the numbers aboveChoose one, choose the other
Choose the L4 if…
- your model needs more than 16 GB per GPU: 24 GB against 16 GB.
Choose the T4 if…
- the hourly price matters most: median $0.49 against $0.80, 39% less.
- you serve LLMs: memory bandwidth per dollar is better ($1.540 per TB/s-hour against $2.683).
- you want choice: 26 providers have a fresh offer today against 25.
Questions about L4 vs T4
Which is cheaper to rent, the L4 or the T4?
Today the cheapest current on-demand L4 is $0.27 per GPU-hour and the cheapest T4 is $0.20; medians are $0.80 and $0.49. Cheapest excludes offers marked unavailable and counts one floor per marketplace.
How much memory do the L4 and T4 have?
24 GB of GDDR6 on the L4, 16 GB of GDDR6 on the T4, at 0.30 and 0.32 TB/s.
Where do these numbers come from?
From 159 current offers read on the providers' own pages and APIs over the last 45 days (reference 00:32 UTC), or in our October 2026 market study for the few providers our collector does not read. Specifications come from manufacturer datasheets. Every price on the GPU pages links to its receipt and shows its read date.