$0
For learning, prototypes, and small projects.
- Up to 100 projects
- 10 branches per project
- Autoscaling up to 2 CU
Neon Pricing, Features & Alternatives
Postgres backend with branching databases, authentication, functions, object storage, and scale-to-zero compute.
Neon separates PostgreSQL compute from storage. Copy-on-write branches reuse existing data, and compute can suspend when idle. That model lets temporary databases avoid full storage copies and idle compute. Neon high availability protects against failures within a selected region; it does not currently provide realtime replication between regions. Failover can restart or reschedule compute against durable storage; it is not an uninterrupted warm-standby compute model. Autoscaling, suspension, recovery history, and replica cost still depend on plan and configuration, and the automated-backup capability refers to built-in retained recovery history rather than independent exported backups.
Best for
Teams that need isolated PostgreSQL databases for previews, CI runs, or coding agents and want those branches to share storage and suspend when idle.
Watch out for
Neon separates storage and compute, so an idle database can be inexpensive while an always-running or heavily branched workload accumulates compute, storage, branch, networking, and support charges.
- Workloads
- APIs
- Real-Time Applications
- Services
- Authentication
- Managed Databases
- Functions
- Object Storage
- Delivery
- Managed
Development & Execution
Supported platform capabilities, execution environments, development technologies, deployment models, integrations, and triggers.
Database support
- PostgreSQL
Capabilities
- Compute Autoscaling
- Storage Autoscaling
- Scale to Zero
- Automatic Failover
- Database Branching
- Point-in-Time Recovery
- Read Replicas
- Automated Backups
Client languages
- C#
- Go
- Java
- JavaScript
- Kotlin
- PHP
- Python
- Ruby
- Rust
- TypeScript
Framework compatibility
- Astro
- Django
- Express.js
- Fastify
- Laravel
- Next.js
- Nuxt
- Remix
- Ruby on Rails
- Spring Boot
- SvelteKit
Integrations
- GitHub
- Terraform
Deployment model
- Serverless (Managed Runtime)
Pricing
Published pricing structure, representative costs, and material billing factors.
- Primary Billing
- Plan based
- Free Tier
- Recurring free allowances
- Commitment
- Launch and Scale have no monthly minimum.
Neon Plans
Managed Postgres plans with metered compute and storage.
Usage-based
Usage-based plan with no monthly minimum.
- Typical spend: $15 per month
- Autoscaling up to 16 CU
- Up to 7-day point-in-time restore
Usage-based
Usage-based plan with expanded limits and compliance features.
- Typical spend: $701 per month
- Autoscaling up to 16 CU or fixed sizes up to 56 CU
- Up to 30-day point-in-time restore
Neon Usage Additional Usage Cost
Compute, storage, branches, and public egress are metered separately.
| Resource | Billing Unit | Price |
|---|---|---|
| Compute (Launch) | per CU-hour | $0.106 |
| Compute (Scale) | per CU-hour | $0.222 |
| Storage | per GB-month | $0.35 |
| Extra branches | per branch-month | $1.50 |
| Public egress | per GB beyond 500 GB included | $0.10 |
Free Tier
No card requiredRecurring allowances included with the Free plan.
- Storage: 0.5 GB per project
- Compute: 100 CU-hours per project per month
Enterprise Pricing
Custom pricing is available with committed-use discounts, dedicated support, custom limits, and custom SLAs.
Billing Commitment
Launch and Scale have no monthly minimum.
Infrastructure & Availability
Where workloads run, how deployment is scoped, and what underlying infrastructure the platform depends on.
Neon currently offers project regions on AWS. Azure region support was deprecated in April 2026 and is not included in this active-region inventory.
Execution
- Single Region
Deployment scope
- Regional
Geographic availability
- North America
- 3
- Latin America
- 1
- Europe
- 2
- Asia Pacific
- 2
Infrastructure providers
- Amazon Web Services
Available regions
| Location | Area | Region ID |
|---|---|---|
| AWS US East (N. Virginia) | North America | aws-us-east-1 |
| AWS US East (Ohio) | North America | aws-us-east-2 |
| AWS US West (Oregon) | North America | aws-us-west-2 |
| AWS South America (São Paulo) | Latin America | aws-sa-east-1 |
| AWS Europe (Frankfurt) | Europe | aws-eu-central-1 |
| AWS Europe (London) | Europe | aws-eu-west-2 |
| AWS Asia Pacific (Singapore) | Asia Pacific | aws-ap-southeast-1 |
| AWS Asia Pacific (Sydney) | Asia Pacific | aws-ap-southeast-2 |
Availability reviewed · Availability source
Resource & Execution Limits
Pooled connections
10,000
Per endpoint
Databases
500
Per branch
Roles
500
Per branch
Security & Compliance
Security Controls
- Access Control
- Organization and project roles
- Multi-Factor Authentication (MFA)
- Account 2FA and organization-wide enforcement
- Encryption
- TLS 1.2+ in transit; AES-256 at rest
- Network Isolation
- Logical customer-data segregation
- IP Allow List
- Database connection restrictions
- Private Networking
- AWS PrivateLink Scale plan
Audits & Certifications
- SOC Reports
- SOC 2 Type II
- ISO Certifications
- ISO/IEC 27001:2022 & 27701:2019
Compliance
- HIPAA
- HIPAA capability Scale plan; BAA required
FAQ
How does Neon database branching work?
A Neon branch is an isolated copy-on-write database environment created from a parent branch. It starts from the parent’s schema and data without duplicating the whole database, then stores its own changes and runs on an independent compute endpoint.
Do I need the Neon serverless driver for a normal Postgres application?
No. Neon works with standard Postgres drivers when the runtime can open normal Postgres connections. The Neon serverless driver is designed for JavaScript and TypeScript environments where raw TCP is unavailable or expensive, including V8-isolate and edge runtimes, and carries queries over HTTP or WebSockets.
How much does Neon cost?
Neon has a $0 Free plan plus usage-based Launch and Scale plans. Paid usage is driven mainly by compute-unit hours and storage, with no monthly minimum for Launch or Scale.
What are the Neon free tier limits?
The Neon Free plan includes 0.5 GB of storage per project and 100 compute-unit hours per project each month. Other plan limits and feature restrictions also apply.
What is a Neon compute unit (CU) and how is it billed?
A Neon compute unit is a normalized compute measure: 1 CU corresponds to 1 vCPU with 4 GB of RAM. Paid compute is billed in CU-hours, so cost follows both the compute size used and how long it is active.
Does Neon require a credit card for the free tier?
No. Neon’s Free plan does not require a credit card. Billing details are needed when you move to paid usage.
Is Neon HIPAA compliant?
Neon supports HIPAA use on its Scale plan when a Business Associate Agreement is in place. Customers still need to configure and use their application and database environment according to their HIPAA responsibilities.
