Explore our credit programs for startups and researchers.
RunPod works with Startups, Academic Institutions, and Enterprises.
1
Develop
Globally distributed GPU
cloud for your AI workloads
cloud for your AI workloads
Deploy any GPU workload seamlessly, so you can focus less on
infrastructure and more on running ML models.
infrastructure and more on running ML models.
PyTorch
ID: twnw98clgxxf2z
$2.89/hour
200 GB Disk: 200 GB Pod Volume
Volume Path: /workspace
1 x H100 PCIe
9 vCPU 50 GB RAM
CA
8654 Mbps
938 Mbps
963 MBps
0
1970-01-01T00:00:00.000Z
create pod network
1
1970-01-01T00:00:01.000Z
create 20GB network volume
2
1970-01-01T00:00:02.000Z
create container runpod/pytorch:3.10-2.0.0-117
3
1970-01-01T00:00:03.000Z
3.10-2.0.0-117 Pulling from runpod/pytorch
4
1970-01-01T00:00:04.000Z
Digest: sha256:2dbf81dd888d383620a486f83ad2ff47540c6cb5e02a61e74b8db03a715488d6
5
1970-01-01T00:00:05.000Z
Status: Image is up to date for runpod/pytorch:3.10-2.0.0-117
6
1970-01-01T00:00:06.000Z
start container
Spin up a GPU pod in seconds
it's a pain to having to wait upwards of 10 minutes for your pods to spin up - we've cut the cold-boot time down to milliseconds, so you can start building within seconds of deploying your pods.
Choose from 50+ templates ready out-of-the-box, or bring your own custom container.
Get setup instantly with PyTorch, Tensorflow, or any other preconfigured environment you might need for your machine learning workflow.
Along with managed and community templates, we also let you configure your own template to fit your deployment needs.
Along with managed and community templates, we also let you configure your own template to fit your deployment needs.
Powerful & Cost-Effective GPUs
for Every Workload
See all GPUsThousands of GPUs across 30+ Regions
Deploy any container on Secure Cloud. Public and private image repos are supported. Configure your environment the way you want.
Zero fees for ingress/egress
Global interoperability
99.99% Uptime
$0.05/GB/month Network Storage
Starting from $2.49/hr
MI300X
192GB VRAM
283GB RAM
24 vCPUs
$2.49/hr
Secure Cloud
Starting from $1.99/hr
H100 PCIe
80GB VRAM
188GB RAM
16 vCPUs
$2.39/hr
Secure Cloud
$1.99/hr
Community Cloud
Starting from $1.19/hr
A100 PCIe
80GB VRAM
117GB RAM
8 vCPUs
$1.64/hr
Secure Cloud
$1.19/hr
Community Cloud
Starting from $1.89/hr
A100 SXM
80GB VRAM
125GB RAM
16 vCPUs
$1.89/hr
Secure Cloud
Starting from $0.4/hr
A40
48GB VRAM
48GB RAM
9 vCPUs
$0.40/hr
Secure Cloud
Starting from $0.69/hr
L40
48GB VRAM
94GB RAM
8 vCPUs
$0.99/hr
Secure Cloud
$0.69/hr
Community Cloud
Starting from $0.79/hr
L40S
48GB VRAM
94GB RAM
16 vCPUs
$0.86/hr
Secure Cloud
$0.79/hr
Community Cloud
Starting from $0.33/hr
RTX A6000
48GB VRAM
50GB RAM
8 vCPUs
$0.49/hr
Secure Cloud
$0.33/hr
Community Cloud
Starting from $0.16/hr
RTX A5000
24GB VRAM
25GB RAM
3 vCPUs
$0.26/hr
Secure Cloud
$0.16/hr
Community Cloud
Starting from $0.34/hr
RTX 4090
24GB VRAM
29GB RAM
6 vCPUs
$0.69/hr
Secure Cloud
$0.34/hr
Community Cloud
Starting from $0.22/hr
RTX 3090
24GB VRAM
24GB RAM
5 vCPUs
$0.43/hr
Secure Cloud
$0.22/hr
Community Cloud
Starting from $0.20/hr
RTX A4000 Ada
20GB VRAM
47GB RAM
4 vCPUs
$0.34/hr
Secure Cloud
$0.20/hr
Community Cloud
2
Scale
Scale ML inference
with Serverless
with Serverless
Run your AI models with autoscaling, job queueing and
sub 250ms cold start time.
Deploy Nowsub 250ms cold start time.
Autoscale in seconds
Respond to user demand in real time with GPU workers that
scale from 0 to 100s in seconds.
scale from 0 to 100s in seconds.
Flex
Workers
Active
Workers
10 GPUs
6:24AM
100 GPUs
11:34AM
20 GPUs
1:34PM
Usage Analytics
Real-time usage analytics for your endpoint with metrics on completed and failed requests. Useful for endpoints that have fluctuating usage profiles throughout the day.
See the console Active
Requests
Completed:
2,277
Retried:
21
Failed:
9
Execution Time
Total:
1,420s
P70:
8s
P90:
19s
P98:
22s
Execution Time Analytics
Debug your endpoints with detailed metrics on execution time. Useful for hosting models that have varying execution times, like large language models. You can also monitor delay time, cold start time, cold start count, GPU utilization, and more.
See the console Real-Time Logs
Get descriptive, real-time logs to show you exactly what's happening across your active and flex GPU workers at all times.
See the console worker logs -- zsh
2024-03-15T19:56:00.8264895Z INFO | Started job db7c79
2024-03-15T19:56:03.2667597Z
0% | | 0/28 [00:00<?, ?it/s]
12% |██ | 4/28 [00:00<00:01, 12.06it/s]
38% |████ | 12/28 [00:00<00:01, 12.14it/s]
77% |████████ | 22/28 [00:01<00:00, 12.14it/s]
100% |██████████| 28/28 [00:02<00:00, 12.13it/s]
2024-03-15T19:56:04.7438407Z INFO | Completed job db7c79 in 2.9s
2024-03-15T19:57:00.8264895Z INFO | Started job ea1r14
2024-03-15T19:57:03.2667597Z
0% | | 0/28 [00:00<?, ?it/s]
15% |██ | 4/28 [00:00<00:01, 12.06it/s]
41% |████ | 12/28 [00:00<00:01, 12.14it/s]
80% |████████ | 22/28 [00:01<00:00, 12.14it/s]
100% |██████████| 28/28 [00:02<00:00, 12.13it/s]
2024-03-15T19:57:04.7438407Z INFO | Completed job ea1r14 in 2.9s
2024-03-15T19:58:00.8264895Z INFO | Started job gn3a25
2024-03-15T19:58:03.2667597Z
0% | | 0/28 [00:00<?, ?it/s]
18% |██ | 4/28 [00:00<00:01, 12.06it/s]
44% |████ | 12/28 [00:00<00:01, 12.14it/s]
83% |████████ | 22/28 [00:01<00:00, 12.14it/s]
100% |██████████| 28/28 [00:02<00:00, 12.13it/s]
2024-03-15T19:58:04.7438407Z INFO | Completed job gn3a25 in 2.9s
2024-03-15T19:56:03.2667597Z
0% | | 0/28 [00:00<?, ?it/s]
12% |██ | 4/28 [00:00<00:01, 12.06it/s]
38% |████ | 12/28 [00:00<00:01, 12.14it/s]
77% |████████ | 22/28 [00:01<00:00, 12.14it/s]
100% |██████████| 28/28 [00:02<00:00, 12.13it/s]
2024-03-15T19:56:04.7438407Z INFO | Completed job db7c79 in 2.9s
2024-03-15T19:57:00.8264895Z INFO | Started job ea1r14
2024-03-15T19:57:03.2667597Z
0% | | 0/28 [00:00<?, ?it/s]
15% |██ | 4/28 [00:00<00:01, 12.06it/s]
41% |████ | 12/28 [00:00<00:01, 12.14it/s]
80% |████████ | 22/28 [00:01<00:00, 12.14it/s]
100% |██████████| 28/28 [00:02<00:00, 12.13it/s]
2024-03-15T19:57:04.7438407Z INFO | Completed job ea1r14 in 2.9s
2024-03-15T19:58:00.8264895Z INFO | Started job gn3a25
2024-03-15T19:58:03.2667597Z
0% | | 0/28 [00:00<?, ?it/s]
18% |██ | 4/28 [00:00<00:01, 12.06it/s]
44% |████ | 12/28 [00:00<00:01, 12.14it/s]
83% |████████ | 22/28 [00:01<00:00, 12.14it/s]
100% |██████████| 28/28 [00:02<00:00, 12.13it/s]
2024-03-15T19:58:04.7438407Z INFO | Completed job gn3a25 in 2.9s
Everything your app needs. All in
one cloud.
99.99%
guaranteed uptime
10PB+
network storage
7,059,116,746
requests
AI Inference
We handle millions of inference requests a day. Scale your machine learning inference while keeping costs low with RunPod serverless.
AI Training
Run machine learning training tasks that can take up to 7 days. Train on our available NVIDIA H100s and A100s or reserve AMD MI300Xs and AMD MI250s a year in advance.
Autoscale
Serverless GPU workers scale from 0 to n with 8+ regions distributed globally. You only pay when your endpoint receives and processes a request.
Bring Your Own Container
Deploy any container on our AI cloud. Public and private image repositories are supported. Configure your environment the way you want.
Zero Ops Overhead
RunPod handles all the operational aspects of your infrastructure from deploying to scaling. You bring the models, let us handle the ML infra.
Network Storage
Serverless workers can access network storage volume backed by NVMe SSD with up to 100Gbps network throughput. 100TB+ storage size is supported, contact us if you need 1PB+.
Easy-to-use CLI
Use our CLI tool to automatically hot reload local changes while developing, and deploy on Serverless when you’re done tinkering.
Secure & Compliant
RunPod AI Cloud is built on enterprise-grade GPUs with world-class compliance and security to best serve your machine learning models.
Lightning Fast Cold-Start
With Flashboot, watch your cold-starts drop to sub 250 milliseconds. No more waiting for GPUs to warm up when usage is unpredictable.
Certifications
RunPod has obtained SOC2 Type 1 Certification as of February 2025. Our data center partners maintain leading compliance standards (including HIPAA, SOC2, and ISO 27001)
Launch your AI application in minutes
Start building with the most cost-effective platform for developing and scaling machine learning models.