[metadata]
application-name: Koyeb
description: Only pay for what you use, by the second. Scale with your team and stay in control with autoscaling limits.
og:description: Only pay for what you use, by the second. Scale with your team and stay in control with autoscaling limits.
og:image: https://www.koyeb.com/og?text=High-performance%20Infrastructure%20for%20APIs,%20Inference,%20and%20Databases
og:locale: en_US
og:site_name: Koyeb
og:title: Pricing for intensive infractrucure - Koyeb
og:type: website
og:url: https://www.koyeb.com/pricing
twitter:card: summary_large_image
twitter:description: Only pay for what you use, by the second. Scale with your team and stay in control with autoscaling limits.
twitter:image: https://www.koyeb.com/og?text=High-performance%20Infrastructure%20for%20APIs,%20Inference,%20and%20Databases
twitter:site: @gokoyeb
twitter:title: Pricing for intensive infractrucure - Koyeb
viewport: width=device-width, initial-scale=1

[canonical-links]
https://www.koyeb.com/pricing

[document-links]
/
All systems operational: https://status.koyeb.com
Blog: /blog
Changelog: /changelog
Contact us: https://tally.so/r/nGLjGo
Contact us: https://tally.so/r/nGLjGo?utm_medium=www&utm_source=koyeb&utm_campaign=pricing_page
Deploy: /deploy
Docs: /docs
Get started: https://app.koyeb.com/auth/signup
Koyeb is joining Mistral AI to Build The Future of AI Infrastructure. Read More →: /blog/koyeb-is-joining-mistral-ai-to-build-the-future-of-ai-infrastructure
Learn more: /blog/koyeb-for-startups-accelerate-with-credits-for-high-performance-infrastructure
Login: https://app.koyeb.com/auth/signin
Partners: /partners
Pricing: /pricing
Request access: https://tally.so/r/npRak8
Sign up: https://app.koyeb.com/auth/signup
Tutorials: /tutorials
View pricing calculator: /calculator
api: /docs/api
blog: /blog
careers: /careers
changelog: /changelog
community: https://community.koyeb.com
company: /about
customer stories: /customers
deploy: /deploy
documentation: /docs
events: https://lu.ma/koyeb
https://slack.koyeb.com
https://www.linkedin.com/company/koyeb
https://x.com/gokoyeb
overview: /
partners: /partners
pricing: /pricing
privacy policy: /docs/legal/data-processing-agreement
startup program: /blog/koyeb-for-startups-accelerate-with-credits-for-high-performance-infrastructure
system status: https://status.koyeb.com
terms of service: /docs/legal/terms
tutorials: /tutorials

[content]
Pricing for intensive infractrucure - Koyeb
Koyeb is joining Mistral AI to Build The Future of AI Infrastructure. Read More →
Pricing
Docs
Blog
Tutorials
Changelog
Deploy
Partners
Login
Sign up
Pricing for intensive infrastructure
Only pay for what you use, by the second. Scale with your team and stay in control with autoscaling limits.
Compute
Postgres
Serverless Compute
We automatically deploy your app in autoscaling, secure VMs on top of high-performance bare metal servers.
/Popular instances
GPU
Standard
Eco
RTX-A6000
vRAM
48GB
vRAM
48GB
$
0.75
/hour
$
0.000208
/sec
L40S
vRAM
48GB
vRAM
48GB
$
1.20
/hour
$
0.000333
/sec
A100
vRAM
80GB
vRAM
80GB
$
1.60
/hour
$
0.000444
/sec
Autoscaling
Scale-to-Zero
Per second billing
Get started
Trusted by the most ambitious teams
Plans
Compute
Postgres
Features
Q&A
Designed for Growth
Pay only for what you use. Add support plans and capacity.
Pro
$
29
/mo
+compute
The best way to start, including team and support features for production apps.
Features
$10 included compute
10 users
100 services
NVMe Volumes and Snapshots
5 concurrent builds
E-mail support and chat
Get started
Scale
$
299
/mo
+compute
Scale your infrastructure, team and support.
Features
$100 included compute
50 users
1000 services
AWS Regions
Slack cross-connect
99.9% uptime SLA
Get started
Enterprise
Custom
+compute
Supercharged security and support features, starting at $1000/mo.
Features
Private dedicated locations
Unlimited users
SSO, RBAC, and Audit trail
Custom RAM, CPU, and GPU
ISO27001 and SOC2 Certifications
99.99% uptime SLA
24×7×365 premium support
Contact us
Early stage startup
Get up to $30k in credits to accelerate your go-to-market with high-performance cloud infrastructure.
Learn more
Saving plans
Long-term reservations get up to 50% off the on-demand price.
Contact us
View pricing calculator
Serverless Compute
When you deploy an app, we automatically deploy inside of secure VMs running on high-performance bare metal servers. Activate autoscaling and we will scale your apps dynamically.
Price per
Month
Hour
Second
Use GPU to deploy:
Autoscaling inference endpoints
High-Performance fine-tuning
Production environments
Using a GitHub repository or pre-built container
Price per
Month
Hour
Second
GPU
Standard
Eco
RTX-4000-SFF-ADA
$
0.50
/hr
vCPU
6
RAM
44GB
Disk
240GB
L4
$
0.70
/hr
vCPU
6
RAM
32GB
Disk
160GB
RTX-A6000
$
0.75
/hr
vCPU
6
RAM
64GB
Disk
160GB
L40S
$
1.20
/hr
vCPU
15
RAM
64GB
Disk
160GB
A100
Popular
$
1.60
/hr
vCPU
15
RAM
180GB
Disk
320GB
A100 SXM
$
2.15
/hr
vCPU
15
RAM
180GB
Disk
320GB
RTX-PRO-6000
$
2.20
/hr
vCPU
15
RAM
220GB
Disk
240GB
H100
$
2.50
/hr
vCPU
15
RAM
180GB
Disk
320GB
H200
$
3.00
/hr
vCPU
15
RAM
240GB
Disk
320GB
2x A100
$
3.2
/hr
vCPU
30
RAM
360GB
Disk
640GB
2x A100 SXM
$
4.30
/hr
vCPU
30
RAM
360GB
Disk
640GB
2x H100
$
5.0
/hr
vCPU
30
RAM
360GB
Disk
640GB
B200
$
5.50
/hr
vCPU
15
RAM
240GB
Disk
320GB
2x H200
$
6.0
/hr
vCPU
30
RAM
480GB
Disk
640GB
4x A100
$
6.4
/hr
vCPU
60
RAM
720GB
Disk
1280GB
4x A100 SXM
$
8.6
/hr
vCPU
60
RAM
720GB
Disk
1280GB
4x H100
$
10.0
/hr
vCPU
60
RAM
720GB
Disk
1280GB
4x H200
$
12.0
/hr
vCPU
60
RAM
960GB
Disk
1280GB
8x A100
$
12.8
/hr
vCPU
120
RAM
1440GB
Disk
2560GB
8x A100 SXM
$
17.2
/hr
vCPU
120
RAM
1440GB
Disk
2560GB
8x H100
$
20.0
/hr
vCPU
120
RAM
1440GB
Disk
2560GB
8x H200
$
24.0
/hr
vCPU
120
RAM
1920GB
Disk
2560GB
N300s
Request access
vCPU
4
RAM
32GB
Disk
320GB
4x N300s
Request access
vCPU
16
RAM
128GB
Disk
320GB
Instance type
vCPU
RAM
Disk
Price
RTX-4000-SFF-ADA
20GB
VRAM
6
44GB
240GB
$
0.50
/hr
L4
24GB
VRAM
6
32GB
160GB
$
0.70
/hr
RTX-A6000
48GB
VRAM
6
64GB
160GB
$
0.75
/hr
L40S
48GB
VRAM
15
64GB
160GB
$
1.20
/hr
A100
Popular
80GB
VRAM
15
180GB
320GB
$
1.60
/hr
A100 SXM
80GB
VRAM
15
180GB
320GB
$
2.15
/hr
RTX-PRO-6000
96GB
VRAM
15
220GB
240GB
$
2.20
/hr
H100
80GB
VRAM
15
180GB
320GB
$
2.50
/hr
H200
141GB
VRAM
15
240GB
320GB
$
3.00
/hr
2x A100
160GB
VRAM
30
360GB
640GB
$
3.2
/hr
2x A100 SXM
160GB
VRAM
30
360GB
640GB
$
4.30
/hr
2x H100
160GB
VRAM
30
360GB
640GB
$
5.0
/hr
B200
180GB
VRAM
15
240GB
320GB
$
5.50
/hr
2x H200
282GB
VRAM
30
480GB
640GB
$
6.0
/hr
4x A100
320GB
VRAM
60
720GB
1280GB
$
6.4
/hr
4x A100 SXM
320GB
VRAM
60
720GB
1280GB
$
8.6
/hr
4x H100
320GB
VRAM
60
720GB
1280GB
$
10.0
/hr
4x H200
564GB
VRAM
60
960GB
1280GB
$
12.0
/hr
8x A100
640GB
VRAM
120
1440GB
2560GB
$
12.8
/hr
8x A100 SXM
640GB
VRAM
120
1440GB
2560GB
$
17.2
/hr
8x H100
640GB
VRAM
120
1440GB
2560GB
$
20.0
/hr
8x H200
1128GB
VRAM
120
1920GB
2560GB
$
24.0
/hr
N300s
24GB
VRAM
4
32GB
320GB
Request access
4x N300s
96GB
VRAM
16
128GB
320GB
Request access
Looking for committed capacity? Let's chat
Serverless Postgres
Fully managed, highly available, and scalable serverless Postgres databases billed by the second.
Storage:
$0.50/mo
per GB
Always active (24/7)
Dev/Staging/CI environments (4h/day)
Business hours (8h/day)
Key features
100% Postgres compatible
Available in Washington, D.C. (US), Frankfurt (Germany), or Singapore (Asia)
Scale-to-zero in case of inactivity
Storage:
$0.50/mo
per GB
Free 5h
$
0
/hr
$
0
/mo
vCPU
0.25
RAM
1 GB
Storage
1GB
Small
$
0.04
/hr
$
29.76
/mo
vCPU
0.25
RAM
1 GB
Storage
∞
Medium
Popular
$
0.08
/hr
$
59.52
/mo
vCPU
0.5
RAM
2 GB
Storage
∞
Large
$
0.16
/hr
$
119.04
/mo
vCPU
1
RAM
4 GB
Storage
∞
XLarge
$
0.32
/hr
$
238.08
/mo
vCPU
2
RAM
8 GB
Storage
∞
2XLarge
$
0.64
/hr
$
476.16
/mo
vCPU
4
RAM
16 GB
Storage
∞
3XLarge
$
1.28
/hr
$
952.32
/mo
vCPU
8
RAM
32 GB
Storage
∞
Instance type
vCPU
RAM
Storage
Price
Free 5h
0.25
1 GB
1GB
$
0
/hr
$
0
/mo
Small
0.25
1 GB
∞
$
0.04
/hr
$
29.76
/mo
Medium
Popular
0.5
2 GB
∞
$
0.08
/hr
$
59.52
/mo
Large
1
4 GB
∞
$
0.16
/hr
$
119.04
/mo
XLarge
2
8 GB
∞
$
0.32
/hr
$
238.08
/mo
2XLarge
4
16 GB
∞
$
0.64
/hr
$
476.16
/mo
3XLarge
8
32 GB
∞
$
1.28
/hr
$
952.32
/mo
Features
Start on-demand and scale with us.
Pro plan
Scale plan
Enterprise plan
Pro
$
29
/mo
+compute
Scale
$
299
/mo
+compute
Enterprise
Custom
+compute
Core
User
10
10
50
Unlimited
Included Usage
$10
$10
$100
$500
Web Services and APIs
Workers
CPU concurrency
100
100
1000
Custom
GPU concurrency
20
20
40
Custom
High-performance bare metal
Horizontal scaling
Autoscaling
Infrastructure autohealing
Serverless Postgres
Volumes and Snapshots
Deploy containers from any registry
Zero-downtime deploys
CLI, API, Terraform, and Pulumi
Build
Deploy on GitHub push
Native build of core runtimes
Build from Docker
Automatic CI/CD
Concurrent builds
5
5
20
Custom
Global deployment
Available regions
7 regions in US, EU, and Asia
7 regions in US, EU, and Asia
7 regions in US, EU, and Asia
50 on-demand regions
AWS Regions
Bring Your Own Cloud (BYOC)
Custom Regions
Dedicated Racks
Networking
Included bandwidth
1TB/mo
1TB/mo
1TB/mo
Custom
Additional bandwidth EU/US
$0.02/GB
$0.02/GB
$0.02/GB
Custom
Additional bandwidth Asia
$0.04/GB
$0.04/GB
$0.04/GB
Custom
Automatic HTTPS/TLS
WebSockets
HTTP/2 and HTTP/3
Private networking
Service discovery
Custom domains
100
100
Unlimited
500 included
Unlimited
500 included
Wildcard domains
Global Load-Balancing
Advanced networking with direct IP
Observability
Realtime metrics
7 days
7 days
7 days
7 days
Realtime logs
7 days
7 days
7 days
30 days
Logs Exporter
Audit logs
Security and Compliance
DDoS Protection
Encryption at rest and in transit
Secrets management
Single Sign-On (SSO)
Role-based Access Control (RBAC)
ISO27001 and SOC2 Certifications
Support
Community Platform
Email
Uptime SLA
99.9%
99.99%
Private Slack support channel
Priority support 24×7×365
Frequently Asked Questions
How are processing and bandwidth charged?
Once you've consumed your included monthly free credit, we start billing you on a pay-per-use basis. Processing is accounted per second and billing is rounded up to the nearest unit. We charge you at the end of the calendar month.
How much do additional Services cost?
Services are charged depending on the effective usage, starting at $0.000006/second. You define the price you want to pay when you choose the Instance size and the scale. If your app requires a $5.40 Instance and you want to launch 10 Instances, you'll pay $54 per month.
What is outbound bandwidth?
All the data we're sending toward end-users through our CDN and outside of our platform is counted as outbound bandwidth.
We're carefully selecting our cloud service providers to get at $0.04/GB transferred. If you have huge volumes to transfer and this doesn't work, contact us to find solutions.
Can I deploy with my own domain?
Yes! You can use your own domain and we automatically provide SSL. We provide ten custom domain for free and each subsequent domain is $0.20/month.
When can I cancel my subscription?
Whenever you want! Your subscription will end immediately and you will receive a prorated credit for the subscription's unused time.
Deploy AI apps to production in minutes
Get started
Talk to an expert
High-performance serverless infrastructure for AI. Run inference, sandboxes, and production workloads on CPUs, GPUs, and accelerators.
All systems operational
Product
overview
pricing
changelog
Resources
documentation
tutorials
community
api
deploy
startup program
system status
Company
blog
customer stories
partners
events
careers
company
terms of service
privacy policy
© Koyeb
