CoreWeave vs Massed Compute
Two GPU specialist providers, priced against each other on the 5 GPUs both of them rent. Published per-GPU on-demand rates as each provider last stated them, with what their own pages say about billing, commitment and egress beside the numbers.
Massed Compute is cheaper on 5 of the 5 cards both of them rent, by a median of 50%.
- How they sell
- GPU specialist An operator whose whole business is renting these cards.
- GPU models
- 10 40 offers
- Registered in
- 🇺🇸 United States
- How they sell
- GPU specialist An operator whose whole business is renting these cards.
- GPU models
- 15 72 offers
- Registered in
- 🇺🇸 United States
The 5 cards both of them rent, cheapest published per-GPU on-demand rate each. Dearest first.
| GPU | CoreWeave | Massed Compute | Difference |
|---|---|---|---|
| NVIDIA B200 SXM6 180GB | $8.60/hr | $5.43/hr | 37% cheaper on Massed Compute |
| NVIDIA H100 SXM5 80GB | $6.16/hr | $2.73/hr | 56% cheaper on Massed Compute |
| NVIDIA A100 80GB | $2.70/hr | $1.35/hr | 50% cheaper on Massed Compute |
| NVIDIA L40S 48GB | $2.25/hr | $0.97/hr | 57% cheaper on Massed Compute |
| NVIDIA L40 48GB | $1.25/hr | $0.86/hr | 31% cheaper on Massed Compute |
Per GPU per hour, on-demand list prices as each provider last published them. Rates exclude tax and vary by region; a cheaper rate is not the same machine, the same network or the same support.
| CoreWeave | Massed Compute | |
|---|---|---|
| Billing granularityOn short jobs this outweighs the rate: an hour minimum on a 90-second run costs more than a 30% dearer per-second rate. | Hourly usage, invoiced monthly in arrears | Per minute |
| Minimum commitment | not stated | None; prepaid balance with $5 minimum threshold |
| Payment model | Invoice on receipt; reserved and bulk credits per order form | Prepaid balance with automatic recharge |
| Free credits | not stated | Referral credit 3% of referred spend; coupon codes |
| Egress | not stated | Free |
| Uptime SLA | 99.9% across regions, 99% single region | 99.999% claimed on product page; no SLA document found |
| Regions | US and European data centres; General and Dedicated Access regions | US regions (e.g. us-central-3); locations not named |
Neither site tells you how fast support answers, whether a commitment auto-renews, how long you have to dispute a bill or whether private networking exists. We looked on both and record the absence rather than filling it in — if you need any of it, ask before you commit.
- NVIDIA H200 SXM5 141GB$6.31/hr
- NVIDIA GH200 96GB$6.50/hr
- NVIDIA GB200 NVL72 192GB$10.50/hr
- NVIDIA A30 24GB$0.35/hr
- NVIDIA RTX A5000 24GB$0.44/hr
- NVIDIA RTX A6000 48GB$0.55/hr
- NVIDIA RTX 6000 Ada 48GB$0.79/hr
- NVIDIA RTX PRO 4500 Blackwell 32GB$0.92/hr
- NVIDIA A100 SXM4 80GB$1.38/hr
- NVIDIA RTX PRO 6000 Blackwell 96GB$2.19/hr
- NVIDIA H100 NVL 94GB$2.92/hr
- NVIDIA H200 NVL 141GB$3.62/hr
- NVIDIA B300 SXM6 288GB$6.60/hr
Is CoreWeave cheaper than Massed Compute?
Massed Compute is cheaper on 5 of the 5 cards both of them rent, by a median of 50%.
Which GPUs do CoreWeave and Massed Compute both rent?
NVIDIA B200 SXM6 180GB, NVIDIA H100 SXM5 80GB, NVIDIA A100 80GB, NVIDIA L40S 48GB, NVIDIA L40 48GB — 5 in total, on published per-GPU on-demand rates.
What is the cheapest GPU at CoreWeave and at Massed Compute?
Of the cards both rent, CoreWeave starts at $1.25 per GPU-hour and Massed Compute at $0.86.