Lambda
CuratedActiveGPU cloud for AI training with on-demand B200/H100 instances and 1-Click Clusters
From
—
GPU from
$0.690/GPU-hr
Locations
4
countries
Data centers
4
sites
Networks (ASN)
2
ASNs
Plans & prices
24
plans
Overview
Lambda, long known as Lambda Labs and a seller of deep-learning workstations, operates a GPU cloud with self-serve on-demand instances, 1-Click Clusters of B200 or H100 GPUs on InfiniBand, and reserved capacity. It moved its main domain from lambdalabs.com to lambda.ai.
Plans & prices
GPU instances24
| Instance | Config | per GPU-hr | Total /hr | Spot | Locations | Source |
|---|---|---|---|---|---|---|
gpu_1x_rtx6000 14 vCPU · 46 GB RAM | 1× Quadro RTX 600024 GB | $0.690 | $0.690 | — |   17 regions | skypilot 3 days ago |
gpu_8x_v100_n 88 vCPU · 448 GB RAM | 8× V100 16GB16 GB | $0.790 | $6.32 | — | 17 regions | skypilot 3 days ago |
gpu_8x_v100 92 vCPU · 448 GB RAM | 8× V100 16GB16 GB | $0.790 | $6.32 | — | 17 regions | skypilot 3 days ago |
gpu_1x_a6000 14 vCPU · 100 GB RAM | 1× RTX A600048 GB | $1.09 | $1.09 | — | 17 regions | skypilot 3 days ago |
gpu_2x_a6000 28 vCPU · 200 GB RAM | 2× RTX A600048 GB | $1.09 | $2.18 | — | 17 regions | skypilot 3 days ago |
gpu_4x_a6000 56 vCPU · 400 GB RAM | 4× RTX A600048 GB | $1.09 | $4.36 | — | 17 regions | skypilot 3 days ago |
gpu_1x_a10 30 vCPU · 200 GB RAM | 1× A1024 GB | $1.29 | $1.29 | — | 17 regions | skypilot 3 days ago |
lambdalabs_A10 30 vCPU · 200 GB RAM | 1× A1024 GB | $1.29 | $1.29 | — | 2 regions | skypilot 2 days ago |
gpu_1x_a100 30 vCPU · 200 GB RAM | 1× A100 40GB40 GB | $1.99 | $1.99 | — | 17 regions | skypilot 3 days ago |
gpu_1x_a100_sxm4 30 vCPU · 200 GB RAM | 1× A100 40GB40 GB | $1.99 | $1.99 | — | 17 regions | skypilot 3 days ago |
lambdalabs_A100-sxm4 30 vCPU · 200 GB RAM | 1× A100 40GB40 GB | $1.99 | $1.99 | — |  3 regions | skypilot 7 hr. ago |
gpu_2x_a100 60 vCPU · 400 GB RAM | 2× A100 40GB40 GB | $1.99 | $3.98 | — | 17 regions | skypilot 3 days ago |
gpu_4x_a100 120 vCPU · 800 GB RAM | 4× A100 40GB40 GB | $1.99 | $7.96 | — | 17 regions | skypilot 3 days ago |
gpu_8x_a100 124 vCPU · 1800 GB RAM | 8× A100 40GB40 GB | $1.99 | $15.92 | — | 17 regions | skypilot 3 days ago |
gpu_1x_gh200 64 vCPU · 432 GB RAM | 1× GH200 (Grace Hopper)96 GB | $2.29 | $2.29 | — | 17 regions | skypilot 3 days ago |
gpu_8x_a100_80gb_sxm4 240 vCPU · 1800 GB RAM | 8× A100 80GB80 GB | $2.79 | $22.32 | — | 17 regions | skypilot 3 days ago |
gpu_1x_h100_pcie 26 vCPU · 200 GB RAM | 1× H100 SXM80 GB | $3.29 | $3.29 | — | 17 regions | skypilot 3 days ago |
gpu_8x_h100_sxm5 208 vCPU · 1800 GB RAM | 8× H100 SXM80 GB | $3.99 | $31.92 | — | 17 regions | skypilot 3 days ago |
gpu_4x_h100_sxm5 104 vCPU · 900 GB RAM | 4× H100 SXM80 GB | $4.09 | $16.36 | — | 17 regions | skypilot 3 days ago |
gpu_2x_h100_sxm5 52 vCPU · 450 GB RAM | 2× H100 SXM80 GB | $4.19 | $8.38 | — | 17 regions | skypilot 3 days ago |
gpu_1x_h100_sxm5 26 vCPU · 225 GB RAM | 1× H100 SXM80 GB | $4.29 | $4.29 | — | 17 regions | skypilot 3 days ago |
gpu_8x_b200_sxm6 208 vCPU · 2900 GB RAM | 8× B200180 GB | $6.69 | $53.52 | — | 17 regions | skypilot 3 days ago |
gpu_2x_b200_sxm6 52 vCPU · 720 GB RAM | 2× B200180 GB | $6.89 | $13.78 | — | 17 regions | skypilot 3 days ago |
gpu_1x_b200_sxm6 26 vCPU · 360 GB RAM | 1× B200180 GB | $6.99 | $6.99 | — | 17 regions | skypilot 3 days ago |
GPU models
Recent changesAll changes →
Thursday, 15 October 2026
1 changes
LambdaWednesday, 14 October 2026
7 changesPlans
Tuesday, 13 October 2026
4 changesData centers · 4
Networks (ASN)
Similar providers
DigitalOceanDeveloper-focused cloud known for simple Droplet VMs, now also selling GPU Droplets
VultrIndependent global cloud with VMs, bare metal and NVIDIA/AMD GPUs in 30+ regions
RunPodDeveloper GPU cloud with per-second Pods, serverless endpoints and community hosts
CoreWeaveLargest independent AI neocloud, renting NVIDIA GB300, B200 and H200 clusters
CrusoeVertically integrated AI cloud building its own power-first data center campuses
Massed ComputeUS GPU cloud renting B200, H200 NVL, H100 and workstation GPUs by the hour
Together AIOpen-model inference API plus self-serve and reserved NVIDIA GPU clusters
Thunder ComputeLow-cost per-minute GPU instances and agent sandboxes for developers