Computable launches a marketplace to buy, sell, and redeem GPU hours by the week with instant liquidity, founded by ex-Jump Trading and Coinbase engineers and backed by Y Combinator.
Hey we are Computable. We spent years building trading infrastructure at Jump Trading and Coinbase. From that point of view, compute looks like energy markets before 2000: everything trades through private bilateral leases, there’s no visible price, and nothing can be resold. The same H100 rents at a 2x spread depending on who’s asking, and once you sign a 24-month lease, it can never change hands.<p>So we built a market for GPU nodes, sold by the calendar week. Here are three things you can do on it that you can’t do anywhere else:<p>- Buy exactly the weeks you need. Three nodes for the last two weeks of October means you pay for those two weeks and walk away when the run is done. You don't have to commit to 6-24 months, and there's no flexibility premium.<p>- Sell back what you don’t use. If your plans change, unused weeks turn back into cash: we keep a market quote posted on every position, so there’s always a price to hit.<p>- Buy (or sell) the future. Lock January in July at a price you set, so the cost of a future run is known before another 40% H100 rate jump makes it not.<p>The first auction is live now at <a href="https://market.getcomputable.com">https://market.getcomputable.com</a>: a block of nodes, August through the end of January. Bidding closes July 31. The bids are sealed. You bid how many nodes, which weeks, and your price. Every clearing price gets published after settlement, because nobody knows what a GPU week is worth and we want the prints to exist in public.<p>Happy to go deep in the comments on anything, including the clearing mechanism, which is a packing problem and was fun to design!
# Computable: Buy GPU hours for the exact weeks you need.
Source: [https://www.getcomputable.com/](https://www.getcomputable.com/)
Computable
Computable is the first and only place where you can easilybuy,sell, andredeemGPU hoursacross calendar weeks,with instant liquidity\.
Rental marketplaces charge extra for that flexibility, but our hours come at wholesale with the freedom to sell what you don't use\.
[Buy GPU hours](https://www.getcomputable.com/#)
THE HOURS YOU NEED, ANY WEEKS ON THE CALENDARINSTANT LIQUIDITY
YBacked by Y CombinatorFounders from Jump Trading and Coinbase
AUGUST THROUGH JANUARY · SOLD BY THE WEEK
## Our first auction
LIMITED CAPACITY
We're auctioning a cluster of GPU nodes, August through next January, sold by the week\.A bid is a flexible shape: how many nodes, which weeks, your price\. Sealed, pay what you bid, and every clearing price gets published after settlement\. Bidding is open now and closes July 31\.
[Join the Auction](https://www.getcomputable.com/#)
SEALED UNTIL CLOSE · JULY 31SAME BID, EARLIER WINS
Full rules and the week\-by\-week calendar are inside once you sign up\.
### Pay for only the weeks you need\.
Pick two nodes for the last two weeks of October, pay for exactly those, and walk away when the run is done\.
### Sell back what you don't use\.
We keep astanding buyback quoteposted on every position, so unused capacity turns back into cash the moment you hit sell\.
### You can buy \(or sell\)the future\.
Lock January in July at a price you set, so you know exactly what the future run costs in case another H100 40% rate jump happens\.
Inferra is building a derivatives exchange for GPU compute, offering perpetual futures for chips like H100 and B200 to enable price discovery and cost hedging, aiming to fix the opaque GPU market.
An analysis of GPU rental prices on Vast.ai reveals that pricing for multi-GPU configurations deviates significantly from per-GPU-hour headline rates due to supply constraints, with larger clusters often being unavailable or more expensive.
Kalshi has created a tool that plots the future price of computing power using prediction markets, aiming to build a forward curve for GPU rental costs.
Expanse is a startup that improves GPU/HPC cluster utilization by predicting job resource needs and providing optimizations, addressing the common problem of over-requesting resources that leads to 30-40% effective utilization.
SkyPilot team open-sources a continuously updated catalog that tracks on-demand and spot pricing for 50 GPU models across 20+ clouds, now browsable online.