compare · gpu marketplace
Salad routes container workloads onto a distributed network of consumer GPUs, targeting the lowest possible batch-inference prices. GPU.ai takes the other approach: one account that prices 12+ clouds on every launch and routes you to the cheapest one with live stock.
at a glance
A fair read of both sides: their genuine strengths, the trade-offs, and what an aggregated market does differently.
Strengths
Trade-offs
The aggregator model
Live per-GPU floor prices across our aggregated fleet. Every row links to the full per-model comparison page.
| GPU | VRAM | GPU.ai / GPU / hr | Stock |
|---|---|---|---|
| RTX A4000 | 16 GB | $0.110/hr | 15 available |
| RTX 3080 | 10 GB | $0.135/hr | 17 available |
| V100 | 16 GB | $0.150/hr | 2 available |
| RTX 3090 | 24 GB | $0.182/hr | 37 available |
| RTX 5080 | 16 GB | $0.226/hr | 25 available |
| RTX 4080 | 16 GB | $0.255/hr | 9 available |
| RTX 4090 | 24 GB | $0.410/hr | 72 available |
| RTX A6000 | 48 GB | $0.450/hr | 141 available |
| L4 | 24 GB | $0.500/hr | 10 available |
| A40 | 48 GB | $0.660/hr | 5 available |
| RTX 5090 | 32 GB | $0.675/hr | 77 available |
| RTX 6000 Ada | 48 GB | $0.730/hr | 32 available |
| A100 40GB | 40 GB | $0.750/hr | 2 available |
| RTX PRO 4500 | 32 GB | $0.810/hr | 4 available |
| L40 | 48 GB | $0.877/hr | 100 available |
| L40S | 48 GB | $0.900/hr | 96 available |
| A100 80GB | 80 GB | $1.32/hr | 18 available |
| RTX PRO 6000 | 96 GB | $1.90/hr | 398 available |
| H100 PCIe | 80 GB | $2.79/hr | 1 available |
| H200 NVL | 141 GB | $3.70/hr | 12 available |
| H200 SXM | 141 GB | $4.04/hr | 48 available |
| H100 SXM | 80 GB | $4.13/hr | 23 available |
| B200 | 192 GB | $8.99/hr | 12 available |
| B300 | 288 GB | $10.11/hr | 8 available |
GPU.ai prices are live per-GPU floor rates from our aggregated fleet, normalized per GPU per hour, and billed per second. We only show models with launchable stock. Last updated 2026-11-16; this page refreshes hourly.
honest verdict
faq
GPU.ai takes a different approach: instead of running one fleet, it prices the whole GPU market on every launch and routes you to the cheapest provider with live stock. If Salad's strengths (among the lowest per-hour gpu prices in the market) are not the thing you're paying for, an aggregated market usually wins on price and availability.
GPU.ai publishes its live per-GPU floor price for every model it carries, currently from $0.110/GPU/hr at the low end. Prices are normalized per GPU per hour and billed per second, and the comparison table on this page is refreshed hourly from live fleet data.
Launches are standard SSH instances with your keys, so most workflows move without changes: pick a GPU, launch, rsync your data, run. There are no commitments to unwind and no minimum spend. You can trial a single instance for a few cents before moving anything.
Honestly: your workload is stateless batch inference that tolerates churn, or if absolute lowest price beats every other requirement. This page's job is a fair comparison, not a hard sell: the live price table above is the argument that matters.