Layer 2.5 · head to head
Vast.ai vs Salad
Two floors of the market — datacentre castoffs against gaming PCs.
| Vast.ai | Salad | |
|---|---|---|
| Can you leave | Movable with effort. Expect to redo images, storage wiring and networking. | Movable with effort. Expect to redo images, storage wiring and networking. |
| Counterparty | You are renting from individuals and small operators. Assume any single machine disappears without notice and design for it. | Your workload runs on gaming PCs belonging to members of the public. This is stated plainly by them and should be stated plainly by you to your security reviewer. |
| What it is | Brokers capacity somebody else owns. Cheapest headline rates, and the machine you get is not the machine you chose. | Brokers capacity somebody else owns. Cheapest headline rates, and the machine you get is not the machine you chose. |
| How you reach it | You push a container image and it runs. No cluster to operate. | You push a container image and it runs. No cluster to operate. |
| Accelerators | Whatever the market is offering — RTX 5090, 4090, PRO 6000, V100, H100 and a long tail | Consumer GPUs only — RTX 3000/4000/5000 series |
| Regions | Global, wherever a host has hardware | Distributed consumer machines |
| Pricing model | Open bidding marketplace, interruptible and on-demand. The clearing price is public without an account. | Per hour, extremely low, interruptible by design. |
| Getting started | Cents. | Very low. |
| Capacity | Deep on consumer silicon, thin and volatile on datacentre parts. | Large in aggregate, individually unreliable. |
| Ownership | Private. | Private. |
Vast.ai
For: Price-sensitive batch work, research, and anyone benchmarking what compute *should* cost before signing a contract.
The catch: Reliability is the price of the price. Hosts vary enormously, verification tiers matter, and a cheap unverified box that dies at hour nine is not cheap.
Economics: Consistently the market floor — reported around $1.49/hr for an H100 where hyperscalers ask $7. Its public API is the closest thing this market has to a spot index, which is why this site measures it.
Salad
For: Massively parallel, interruption-tolerant, non-sensitive batch work. Image generation at volume is the canonical fit.
The catch: No datacentre GPUs, no compliance story, no guarantees. For the right workload none of that matters and the price is unbeatable.
Economics: The lowest cost per unit of consumer-GPU work available anywhere, by a wide margin.
Neither table row is a price
Deliberately. Published on-demand rates at this layer move weekly, and essentially nobody signing a real contract pays them — every serious buyer pays less than every list figure either of these companies publishes. Quoting one here would date this page within a month.
The GPU rental price index carries dated, sourced figures instead, and the durable finding there is the spread: the identical H100 rents from roughly $1.38 to $12.29 an hour depending only on who you rent it from.
The layers underneath both
Whichever you pick is renting you chips in a building that needs power. In 2026 that is the constraint that binds: Microsoft has disclosed an Azure backlog it cannot fill for want of megawatts rather than accelerators, and the US interconnection queue exceeds 2,600 GW with roughly 80% of projects withdrawing before they energise.
Layer 2 — Silicon · Layer 1 — Energy · The interconnection queue