GPU prices · September 2026
Cloud GPU prices per hour
The cheapest on-demand list price per GPU-hour for each GPU, across the neoclouds in the public SkyPilot catalog. Pick a GPU for every provider's price.
Updated (13 minutes ago)
| GPU | Memory | From $/GPU-hr | Cheapest on | Providers |
|---|---|---|---|---|
| H100 | 80 GB HBM3 | $1.29 | Hyperbolic | 6 |
| H200 | 141 GB HBM3e | $2.60 | Seeweb | 4 |
| B200 | 180 GB HBM3e | $6.69 | Lambda | 3 |
| A100 80GB | 80 GB HBM2e | $0.99 | Seeweb | 6 |
| A100 40GB | 40 GB HBM2 | $1.99 | Lambda | 1 |
| L40S | 48 GB GDDR6 | $0.85 | Seeweb | 4 |
| GH200 | 96 GB HBM3 | $2.29 | Lambda | 1 |
| RTX 6000 Ada | 48 GB GDDR6 | $0.84 | RunPod | 2 |
| MI300X | 192 GB HBM3 | $2.39 | RunPod | 1 |
| A6000 | 48 GB GDDR6 | $0.50 | Hyperstack | 4 |
| RTX 4090 | 24 GB GDDR6X | $0.74 | RunPod | 1 |
GPUs listed by fewer than 3 providers don't get their own page: one or two prices aren't a market.
Launch on MatMul
MatMul shows what you can launch right now across neoclouds, cheapest first, and bills it from one prepaid balance. Its prices are the supplier's price plus a markup, so they differ from the list prices on this page.
How we get these prices
Prices come from the public SkyPilot catalog (Apache-2.0), which lists each neocloud's on-demand list prices. We read it every 30 minutes and divide each instance's hourly price by its GPU count, so every figure is per GPU-hour. For each provider we show its cheapest listing; spot and peer-to-peer marketplace listings aren't included.
A list price is not a guarantee: the provider may have no free capacity at that price, and what you can actually launch changes minute to minute. MatMul launches from live supplier inventory and charges the supplier's price plus a markup, so the price you pay on MatMul is not the market floor shown here. The console shows the exact hourly price before you launch.