All cloud GPUs, by rental price
Every one of the 32 GPU models we track, grouped by tier and ranked by on-demand price when available. Non-firm models are labeled by their listed tier instead of being presented as firm capacity. Click any model for its full provider table, price history and rent-vs-buy math.
Datacenter GPUs
Server-class accelerators (H100, A100, MI300X and up) for training and large-scale inference.
| GPU | VRAM | On-demand from | Spot/community from | Providers | Cheapest at | |
|---|---|---|---|---|---|---|
| Tesla V100 | 16 GB | $0.17 | $0.19 | 6 | DataCrunch on-demand | Rent → |
| Tesla T4 | 16 GB | $0.35 | $0.19 | 5 | Google Cloud on-demand | Rent → |
| A30 | 24 GB | $0.35 | – | 1 | Massed Compute on-demand | Rent → |
| L4 | 24 GB | $0.44 | $0.10 | 10 | Jarvislabs on-demand | Rent → |
| A40 | 48 GB | $0.44 | – | 2 | RunPod secure cloud | Rent → |
| L40 | 48 GB | $0.79 | $0.69 | 6 | Thunder Compute on-demand | Rent → |
| L40S | 48 GB | $0.88 | $0.79 | 16 | Massed Compute on-demand | Rent → |
| A100 | 40-80 GB | $0.89 | $0.68 | 21 | Jarvislabs on-demand | Rent → |
| A10 | 24 GB | $1.00 | – | 5 | OVHcloud on-demand | Rent → |
| H100 | 80-94 GB | $1.99 | $0.94 | 32 | Voltage Park on-demand | Rent → |
| GH200 | 96 GB | $2.29 | – | 2 | Lambda on-demand | Rent → |
| MI300X | 192 GB | $2.39 | – | 4 | RunPod secure cloud | Rent → |
| MI355X | 288 GB | spot / interruptible only | $2.59 | 1 | Vultr spot / interruptible | Rent → |
| H200 | 141-143 GB | $2.60 | $2.51 | 22 | GMI Cloud on-demand | Rent → |
| B200 | 180 GB | $3.49 | $3.20 | 18 | Prime Intellect on-demand | Rent → |
| MI325X | 256 GB | $3.80 | $2.00 | 2 | DigitalOcean on-demand | Rent → |
| B300 | 288 GB | $4.99 | $4.70 | 12 | Prime Intellect on-demand | Rent → |
| MI350X | 288 GB | $6.16 | – | 1 | DigitalOcean on-demand | Rent → |
| GB200 | 186 GB | $8.00 | – | 1 | GMI Cloud on-demand | Rent → |
| GB300 | 288 GB | $8.62 | – | 2 | DataCrunch on-demand | Rent → |
Workstation GPUs
Professional RTX cards balancing VRAM and price for rendering, fine-tuning and mid-size models.
| GPU | VRAM | On-demand from | Spot/community from | Providers | Cheapest at | |
|---|---|---|---|---|---|---|
| RTX A4000 | 16 GB | $0.10 | $0.17 | 3 | TensorDock on-demand | Rent → |
| RTX A5000 | 24 GB | $0.27 | $0.16 | 2 | RunPod secure cloud | Rent → |
| RTX A6000 | 48 GB | $0.35 | $0.33 | 8 | Thunder Compute on-demand | Rent → |
| Quadro RTX 5000 | 16 GB | $0.60 | – | 1 | OVHcloud on-demand | Rent → |
| Quadro RTX 6000 | 24 GB | $0.69 | – | 1 | Lambda on-demand | Rent → |
| RTX PRO 4500 | 32 GB | $0.72 | – | 2 | RunPod secure cloud | Rent → |
| RTX 6000 Ada | 48 GB | $0.75 | $0.74 | 5 | TensorDock on-demand | Rent → |
| RTX PRO 6000 | 96 GB | $1.10 | $0.48 | 14 | Google Cloud on-demand | Rent → |
Consumer GPUs
GeForce RTX cards on community marketplaces: the cheapest way to rent for small models and prototyping.
| GPU | VRAM | On-demand from | Spot/community from | Providers | Cheapest at | |
|---|---|---|---|---|---|---|
| RTX 5080 | 16 GB | community cloud only | $0.18 | 2 | SaladCloud community cloud | Rent → |
| RTX 3090 | 24 GB | $0.20 | $0.09 | 3 | TensorDock on-demand | Rent → |
| RTX 4090 | 24 GB | $0.35 | $0.16 | 4 | TensorDock on-demand | Rent → |
| RTX 5090 | 32 GB | $0.86 | $0.25 | 3 | Spheron on-demand | Rent → |
Not currently listed anywhere
We track this card, but no provider we follow is renting it today. The page keeps the specifications and whatever price history we recorded before the listings stopped.
On-demand = fixed-price capacity. Spot/community = interruptible marketplace capacity (serverless per-request pricing is shown on each GPU page). Prices are per GPU per hour before storage and egress, rebuilt from the daily snapshot. See also the interactive screener, provider list and price movers.