GPU Rental Pricing
Lium bills every GPU per second with no minimum and no reservation. 29 of 100 GPU models can be rented right now, from $0.06 per GPU-hour; the others show a reference price until a provider lists one.
Prices and availability as of 02:15 UTC
Refer a friend: earn 5% of their top-ups for 3 months, then 2.5% for 3 more. Get your link
| GPU | $ / GPU / hour | Available now |
|---|---|---|
| B300 SXM6 | $4.00 – $11.90 | 5 pods · 40 GPUs |
| B200 | $3.70 – $8.89 | 5 pods · 29 GPUs |
| H200 | $3.50 – $5.38 | 7 pods · 33 GPUs |
| H200 NVL | $3.65 | 1 pods · 1 GPUs |
| H100 80GB HBM3 | $1.85 – $5.00 | 4 pods · 27 GPUs |
| H100 PCIe | $1.90 | 1 pods · 1 GPUs |
| RTX PRO 6000 Blackwell Server Edition | $1.49 – $2.49 | 8 pods · 46 GPUs |
| RTX PRO 6000 Blackwell Workstation Edition | $1.87 – $2.00 | 4 pods · 11 GPUs |
| RTX 6000 Ada Generation | $0.79 – $1.49 | 8 pods · 36 GPUs |
| RTX PRO 5000 Blackwell | $0.80 | 3 pods · 5 GPUs |
| A100 80GB PCIe | $0.81 | 1 pods · 1 GPUs |
| RTX A6000 | $0.40 – $0.67 | 16 pods · 40 GPUs |
| RTX 5090 | $0.50 – $0.85 | 35 pods · 146 GPUs |
| RTX 4090 | $0.34 – $0.57 | 26 pods · 85 GPUs |
| CMP 170HX | $0.37 | 1 pods · 2 GPUs |
| Graphics Device | $0.50 – $1.00 | 3 pods · 4 GPUs |
| RTX 4070 Ti SUPER | $0.20 | 1 pods · 3 GPUs |
| RTX 4080 SUPER | $0.22 | 1 pods · 1 GPUs |
| RTX 4080 | $0.20 | 1 pods · 1 GPUs |
| RTX 3090 Ti | $0.25 | 1 pods · 1 GPUs |
| RTX 3090 | $0.16 – $0.30 | 16 pods · 31 GPUs |
| RTX 5080 | $0.27 – $0.40 | 2 pods · 3 GPUs |
| RTX 5070 Ti | $0.24 – $0.25 | 7 pods · 21 GPUs |
| RTX 5060 Ti | $0.15 – $0.18 | 5 pods · 19 GPUs |
| RTX 4070 Ti | $0.15 | 1 pods · 1 GPUs |
| RTX 4060 Ti | $0.14 – $0.15 | 3 pods · 5 GPUs |
| RTX 5070 | $0.17 | 1 pods · 2 GPUs |
| GTX 1660 Ti | $0.10 | 1 pods · 1 GPUs |
| RTX 3060 Ti | $0.06 | 1 pods · 3 GPUs |
| GB300 | $8.00 reference | None right now |
| GB200 | $5.60 reference | None right now |
| RTX PRO 6000 Blackwell Max-Q Workstation Edition | $1.19 reference | None right now |
| H100 NVL | $1.11 reference | None right now |
| H800 80GB HBM3 | $0.88 reference | None right now |
| RTX PRO 6000D Blackwell Workstation Edition | $0.82 reference | None right now |
| RTX 6000D | $0.82 reference | None right now |
| H800 NVL | $0.80 reference | None right now |
| H800 PCIe | $0.80 reference | None right now |
| RTX 2080 SUPER | $0.74 reference | None right now |
| A100-SXM4-80GB | $0.70 reference | None right now |
| A100-SXM4-40GB | $0.48 reference | None right now |
| RTX PRO 4500 Blackwell | $0.40 reference | None right now |
| RTX PRO 4500 Blackwell Server Edition | $0.40 reference | None right now |
| L40S | $0.38 reference | None right now |
| A100-PCIE-40GB | $0.37 reference | None right now |
| GTX 1070 | $0.37 reference | None right now |
| RTX 5880 Ada Generation | $0.34 reference | None right now |
| L40 | $0.33 reference | None right now |
| A800 80GB PCIe | $0.31 reference | None right now |
| RTX 5000 Ada Generation | $0.27 reference | None right now |
| GB10 | $0.25 reference | None right now |
| Quadro RTX 8000 | $0.21 reference | None right now |
| RTX 4500 Ada Generation | $0.20 reference | None right now |
| RTX PRO 4000 Blackwell | $0.20 reference | None right now |
| A10 | $0.20 reference | None right now |
| RTX PRO 2000 Blackwell | $0.17 reference | None right now |
| Tesla V100 | $0.17 reference | None right now |
| RTX 4000 Ada Generation | $0.16 reference | None right now |
| RTX A5000 | $0.16 reference | None right now |
| RTX 5090 D | $0.15 reference | None right now |
| T4 | $0.15 reference | None right now |
| RTX A4500 | $0.13 reference | None right now |
| Quadro RTX 6000 | $0.13 reference | None right now |
| RTX 4000 SFF Ada Generation | $0.13 reference | None right now |
| RTX A4000 | $0.12 reference | None right now |
| A40 | $0.12 reference | None right now |
| TITAN RTX | $0.12 reference | None right now |
| RTX 5060 | $0.11 reference | None right now |
| RTX 4090 D | $0.11 reference | None right now |
| L4 | $0.11 reference | None right now |
| A30 | $0.10 reference | None right now |
| TITAN V | $0.10 reference | None right now |
| RTX 3080 Ti | $0.10 reference | None right now |
| RTX 4070 SUPER | $0.09 reference | None right now |
| RTX 3080 | $0.09 reference | None right now |
| RTX 3070 Ti | $0.09 reference | None right now |
| GTX 1060 | $0.09 reference | None right now |
| RTX 4070 | $0.08 reference | None right now |
| RTX 3060 Laptop GPU | $0.08 reference | None right now |
| RTX 2080 Ti | $0.08 reference | None right now |
| Tesla P100 | $0.08 reference | None right now |
| Tesla M40 | $0.08 reference | None right now |
| RTX 2000 Ada Generation | $0.07 reference | None right now |
| RTX 3070 | $0.07 reference | None right now |
| Quadro RTX 5000 | $0.07 reference | None right now |
| RTX 2070 SUPER | $0.07 reference | None right now |
| GTX 1660 | $0.07 reference | None right now |
| Tesla P40 | $0.07 reference | None right now |
| RTX 4060 | $0.06 reference | None right now |
| RTX A2000 | $0.06 reference | None right now |
| RTX 3060 | $0.06 reference | None right now |
| RTX 2060 SUPER | $0.06 reference | None right now |
| RTX 3050 | $0.05 reference | None right now |
| GTX 1660 SUPER | $0.05 reference | None right now |
| Quadro P4000 | $0.05 reference | None right now |
| RTX 2060 | $0.04 reference | None right now |
| TITAN Xp | $0.04 reference | None right now |
| GTX 1080 Ti | $0.04 reference | None right now |
| GTX 1070 Ti | $0.04 reference | None right now |
| GTX 1080 | $0.03 reference | None right now |
Frequently asked questions
What do the prices on this page mean?
One GPU for one hour, set by the provider. A range spans the live asks; with no free pod, it is the reference price. 29 of 100 models have a free pod now, 599 GPUs in total.
Does Lium have per-second billing with no minimum?
Yes. Every pod is billed per second, with no minimum and no reservation. To start one, your balance must cover 15 minutes of it plus 15 minutes of each pod already running, starting or awaiting an automatic retry (60 minutes of each while the account has only the signup credit). Billing stops when you delete the pod or it reaches its scheduled end time.
Can I pay for GPUs with crypto?
Yes. Top up by card, crypto (pick the coin and network in the Top Up form), or TAO or Bittensor alpha. Crypto, TAO and alpha top-ups get a 2.5 % bonus.
Where can a script or an aggregator read these prices?
The table is JSON at lium.io/pricing.json, each model page is served as markdown to a client that asks for text/markdown, and live nodes are at lium.io/api/public/v1/nodes. No account needed.
Do the prices include storage or a platform fee?
The price shown is the full hourly price of the pod, local disk included; the provider receives 95 % of what you pay for the rental. A persistent volume is billed separately, per GB-hour of storage plus data read out, both at cost.
For AI agents (CLI install, agent skill, headless signup): llms-full.txt