Favicon of Neon

Neon Serverless Postgres

Neon helps software companies and developers manage Postgres databases without manual server provisioning. It is designed for teams building AI applications or platforms that require isolated databases per tenant.

At a glance

Best for
Software companies, AI application developers, SaaS platform operators, Engineering teams
Pricing
Neon offers a free tier with 100 projects. Paid plans are usage-based, starting at $0.106 per CU-hour for the Launch plan. A startup program is available offering up to $100k in credits.
Key use cases
AI Agent Development, Database-per-Tenant Architectures, Development and Testing, Serverless Application Backend
Integrations
Datadog, OTel
Official website
neon.com
Screenshot of Neon website

Neon is a serverless Postgres database platform, now part of Databricks, that separates compute and storage to help remove the need for manual infrastructure management. The platform automatically adjusts resources based on query load, which may help reduce costs associated with idle capacity.

It is designed for developers and software companies, particularly those creating AI agents or SaaS platforms. The platform includes an API for fleet management, which supports the programmatic provisioning of isolated databases for users.

Buyers should confirm their requirements for time-travel and recovery, as the restore window and historical data retention vary by plan. Those needing specific compliance and SLAs should review the Scale plan specifications.

Key Features

Autoscaling Compute

Adjusts CPU and memory based on workload demand and can scale down to zero when inactive.

Instant Database Branching

Creates editable copies of databases using copy-on-write storage for development and testing.

Built-in Authentication

Includes user authentication and management features integrated into the database.

Connection Pooling

Uses a pooler based on pgBouncer to support up to 10,000 concurrent connections.

Point-in-Time Recovery

Supports restoring a database to a previous state based on a configurable restore window.

Fleet Management API

Supports programmatic provisioning, quota control, and scaling for large numbers of databases.

Use Cases

AI Agent Development

Providing the backend for AI agents with automatic scaling and integrated authentication.

Database-per-Tenant Architectures

Provisioning isolated Postgres databases for individual users within a SaaS platform via API.

Development and Testing

Using instant branching to create environments for testing without affecting live data.

Serverless Application Backend

Managing database resources for apps with fluctuating traffic patterns via autoscaling.

Best For

Software companiesAI application developersSaaS platform operatorsEngineering teams

Integrations

DatadogOTel

Pricing

Neon offers a free tier with 100 projects. Paid plans are usage-based, starting at $0.106 per CU-hour for the Launch plan. A startup program is available offering up to $100k in credits.

FAQ

What is a Compute Unit (CU) in Neon?

A CU represents the compute size, allocating approximately 1 vCPU and 4 GB of RAM. Billing is based on CU-hours used.

Does Neon have a free tier?

Yes, the Free plan allows for up to 100 projects, each with 100 CU-hours monthly and 0.5 GB of storage.

How does Neon's branching work?

Neon uses copy-on-write storage to create instant, editable copies of a database, which allows for testing changes without duplicating the entire dataset.

Source category: Software Development

Source subcategory: Database

Software Type:

Featured Tools

Favicon
  
  
 
   
Favicon
  
  
 
   
Favicon
  
  
 
   
Favicon
  
  
 
   
Favicon
  
  
 
   
Favicon