/* pricing */

Pay for what you actually use.

Predictable hourly compute. Scale down or shut off and the bill stops the same minute. No included-hours math, no surprise overages, no upsell paywalls hiding standard features.

/* account tiers */

Pick the account that fits your work.

Account tier sets what features your account unlocks. Compute is billed separately, hourly, on every plan.

/* standard */

$0

+ compute usage, billed hourly

  • Auto-deploys from GitHub
  • Unlimited custom domains
  • Zero-downtime deploys
  • 300 build minutes/mo
  • 7 days log retention

/* pro */

$29 /mo

+ compute usage, billed hourly

  • Everything in Standard
  • Dockerfile builds
  • Multiple environments
  • Smart autoscaling
  • 1,000 build minutes/mo
  • 30 days log retention

/* team */

$49 /mo

+ compute usage, billed hourly

  • Everything in Pro
  • 5 seats included (+$5/seat)
  • Member roles
  • Audit logs
  • 3,000 build minutes/mo
  • Priority support

/* compute */

Pick a size. Pay by the hour.

Billed hourly with a minimum charge of one hour. Scale down anytime.

Standard tier — shared CPU

Burstable CPU can use spare capacity when available.

Size vCPU RAM Hourly Monthly
standard-1x Shared 512 MB $0.007 $5
standard-2x Shared 1 GB $0.014 $10
standard-4x Shared 2 GB $0.028 $20

Performance tier — dedicated CPU

Guaranteed resources, never throttled.

Size vCPU RAM Hourly Monthly
performance-1x 1 CPU 2 GB $0.042 $30
performance-2x 2 CPUs 4 GB $0.076 $55
performance-4x 4 CPUs 8 GB $0.153 $110
performance-8x 8 CPUs 16 GB $0.306 $220

/* build minutes */

Only active build time.

No queue waits, no scheduler overhead — we bill what runs.

Plan Included Timeout Concurrent Overage
Standard 300/mo 20 min 1 $0.007/min
Pro 1,000/mo 45 min 2 $0.006/min
Team 3,000/mo 60 min 5 $0.005/min

Most apps stay well within their included pool.

/* bandwidth */

Bandwidth that scales with you.

Included bandwidth scales with compute. Overage is cheap.

Included

20 GB per $1 of compute

Overage

$0.05/GB

Monthly compute Included bandwidth
$5 100 GB
$20 400 GB
$50 1 TB
$100 2 TB

Most apps never exceed included bandwidth.

/* feature comparison */

What's in each plan.

Feature Standard Pro Team
Deployment & Builds
Deploy via Git ✓ Team Emails
Buildpacks
Dockerfile Builds
CI/CD (GitHub)
Processes & Resources
Web Process
Worker Processes
Cron Jobs
One-Off Commands
Environments
Multiple Environments
Preview Environments
Team & Security
Team Management
Seats 5 included (+$5/seat)
Member Roles
Audit Logs
2-Factor Auth ✓ + Enforcement
Application Configuration
Environment Variables App Level App Level App & Team Level
HTTP/2 & HTTP/3
WebSockets
Custom Domains ✓ Unlimited ✓ Unlimited ✓ Unlimited
Wildcard Domains
SSL Certificates ✓ Free ✓ Free ✓ Free
Scaling & Metrics
Application Metrics
Zero-Downtime Deploys
Smart Autoscaling
Clone / Duplicate Apps
Usage & Billing Alerts
Observability
Integrated Logs
Log Retention 7 days 30 days 30 days
Notifications Email Email, Slack, Discord, Telegram Email, Slack, Discord, Telegram
Platform
All Regions
Private Regions $499/mo
API Access
Release Rollbacks 3 10 10
Support Standard Standard Priority

/* faq */

Questions, answered.

How does billing work?

Two parts: your account tier (Standard is $0; Pro is $29/mo; Team is $49/mo) and the compute your apps actually use, billed hourly. Scale down or shut off and the bill stops the same minute.

Is there a free tier?

Accounts are free to create; compute is metered from minute one, starting at around $5/mo for the smallest pod. We're customer-funded, not VC-subsidized — so the math only works if everyone running real CPU pays for it. If a $0 deploy is a hard requirement, you're better served by a platform built around that.

Standard or performance compute — which should I pick?

Standard for web apps, APIs, and bursty work — shared cores that burst when there's spare capacity. Performance for CPU-heavy work that needs guaranteed cycles — dedicated cores, never throttled.

Are there cold starts?

No. Your processes stay running between requests; the first request after a quiet stretch is as fast as the millionth.

Can I upgrade or downgrade anytime?

Yes. Plan changes apply immediately. Compute changes apply on your next deploy or scale operation.

Are there any hidden fees?

No. You pay your plan fee, the compute your apps consume, and bandwidth overages if you blow past your included pool. That's it.

What if I want to leave?

Your repo just works on the next platform that speaks buildpacks — or on a server you set up yourself. No proprietary runtime to unwind, no exit tax, no lock-in.

What payment methods do you accept?

All major credit cards.

/* ready when you are */

Your next deploy is one push away.