>>

HIGH-PERFORMANCE GPU PODS

Give your AI agent a GPU in one command.
Agent quickstart
Per-second billing from RTX 5090 to B300. Your agent starts a pod, runs the job and removes it, within a budget you set.
pip install lium.io && lium up --gpu H100 --ttl 1h# auto-removes after 1h
Showing
9 / Available Pods
8X

NVIDIA B300 SXM6 AC

SPOT
$55.12
/HOUR
$6.89/GPU
RENT NOW
CPU
192X INTEL(R) XEON(R) PLATINUM 8559C
Memory
4,012 GB
VRAM
2,149 GB
Hard disk
6,003 GB
# of ports
5
Network
9.20 Gbps
7.14 Gbps
Uptime
1 DAYS 10 HRS 30 MINS
Location
Incheon, South Korea
8X

NVIDIA B300 SXM6 AC

SPOT
$54.8
/HOUR
$6.85/GPU
RENT NOW
CPU
192X INTEL(R) XEON(R) PLATINUM 8559C
Memory
4,012 GB
VRAM
2,149 GB
Hard disk
6,047 GB
# of ports
5
Network
4.95 Gbps
5.17 Gbps
Uptime
1 DAYS 21 HRS 45 MINS
Location
Jakarta, Indonesia
CPU
224X INTEL(R) XEON(R) 6776P
Memory
1,890 GB
VRAM
2,149 GB
Hard disk
37,328 GB
# of ports
300
Network
3.45 Gbps
3.08 Gbps
Uptime
1 MONTHS 28 DAYS 2 HRS 30 MINS
Reliability
98%
Location
Montreal, Canada
CPU
208X INTEL(R) XEON(R) PLATINUM 8480+
Memory
1,772 GB
VRAM
637 GB
Hard disk
21,750 GB
# of ports
300
Network
6.55 Gbps
2.97 Gbps
Uptime
2 DAYS 12 HRS 15 MINS
Reliability
95%
Location
Austin, United States
CPU
240X AMD EPYC 9575F 64-CORE PROCESSOR
Memory
2,008 GB
VRAM
2,149 GB
Hard disk
1,116 GB
# of ports
300
Network
2.28 Gbps
2.80 Gbps
Uptime
24 DAYS 7 HRS 15 MINS
Reliability
100%
Location
Helsinki, Finland
CPU
208X INTEL(R) XEON(R) PLATINUM 8480+
Memory
1,772 GB
VRAM
637 GB
Hard disk
21,750 GB
# of ports
300
Network
5.77 Gbps
1.85 Gbps
Uptime
2 DAYS 12 HRS 15 MINS
Reliability
95%
Location
Dallas, United States
1X

NVIDIA H200 NVL

GPU Splitting · from 1 GPU
$3.85
/HOUR
RENT NOW
CPU
128X AMD EPYC 9334 32-CORE PROCESSOR
Memory
126 GB
VRAM
140 GB
Hard disk
620 GB
# of ports
289
Network
933.77 Mbps
929.04 Mbps
Uptime
7 DAYS 5 HRS 15 MINS
Reliability
98%
Location
Montreal, Canada
CPU
12X 12TH GEN INTEL(R) CORE(TM) I5-12400F
Memory
63 GB
VRAM
32 GB
Hard disk
745 GB
# of ports
18
Network
378.38 Mbps
708.12 Mbps
Uptime
4 DAYS 20 HRS 15 MINS
Reliability
89%
Location
,
1X

NVIDIA GeForce RTX 3090

GPU Splitting · from 1 GPU
Reduced power limit
$0.29
/HOUR
RENT NOW
CPU
72X INTEL(R) XEON(R) CPU E5-2697 V4 @ 2.30GHZ
Memory
134 GB
VRAM
24 GB
Hard disk
6,022 GB
# of ports
78
Network
94.22 Mbps
117.59 Mbps
Uptime
2 MONTHS 9 DAYS 1 HRS
Reliability
93%
Location
Frankfurt am Main, Germany
For AI agents — CLI install, agent skill, headless signup: llms-full.txt