Changelog

Add organization members by domain

You can now add members to your organization by email domain. Organization admins can add and verify one or more domains (for example, yourcompany.com) in your Neon organization Settings, under Domains.

When a user signs up or logs in to Neon with an email that matches a verified domain, they’re automatically added to your Neon organization as a Member, no invite email required. They see your organization in the org switcher in the Neon Console.

This is useful for teams that want everyone within the company email domain to have access without needing to send individual invites. You simply add your domain in the Neon Console, add a TXT record at your DNS provider to verify ownership, then click Verify in the Neon Console.

Colleagues who already have Neon accounts are added to the organization the next time they log in. For the full flow and behavior (roles, multiple domains), see Add members by domain.

Neon MCP Server updates

This week’s Neon MCP Server release brings new tools for pulling Neon documentation and setup guidance into your development environment, plus a new guide for connecting Google Jules to the Neon MCP Server.

New documentation retrieval tools

The MCP Server now includes two tools so your AI agent or MCP client can fetch Neon docs on demand:

• list_docs_resources – Lists all available Neon documentation pages from the docs index. Returns page URLs and titles so you can choose which page to load.
• get_doc_resource – Fetches a specific Neon documentation page as markdown. Use list_docs_resources first to discover page slugs, then pass the slug to this tool to load the content.

Together, these tools let your agent or assistant look up setup, configuration, and how-to content from the Neon docs without leaving the chat.

Neon MCP Server on Google Jules

The Neon MCP Server is now available in Google Jules, Google's AI-powered coding assistant. Create a Neon API key, add the server in Jules settings, and you're set. Full setup steps are in Connect MCP clients to Neon.

Compute Autoscaling Report

We've published a Compute Autoscaling Report that breaks down how Neon's autoscaling compute compares to provisioned, fixed compute sizes, based on real production workloads that run on Neon.

The report walks through what happens when you use provisioned, fixed compute sizes vs. autoscaling compute, and how that impacts cost and performance. If you've ever wondered how much autoscaling actually saves you (or how it behaves under real traffic), the report lays it out with real data and the full methodology.

To learn more about Neon's autoscaling feature and how to enable it for your projects, see Autoscaling.

Custom API key header for OpenTelemetry

You can now specify a custom header name for API key authentication when configuring OpenTelemetry integrations. The header defaults to X-API-Key if not specified. This makes it easier to integrate with services like Honeycomb that expect a different header name for API keys.

Track your usage programmatically on paid plans

A new consumption history API for current usage-based plans is now available on all paid plans, including Launch. The metrics returned align directly with usage-based billing, so what you query matches what you see on your invoice.

Use this API to build custom dashboards, integrate with your reporting tools, or set up usage alerts. Query at hourly, daily, or monthly granularity for metrics like compute usage, storage (root and child branches), instant restore, data transfer, and extra branches.

This example retrieves month-to-date usage for all metrics:

curl --request GET \
     --url 'https://console.neon.tech/api/v2/consumption_history/v2/projects?from=2026-02-01T00:00:00Z&to=2026-02-06T00:00:00Z&granularity=daily&org_id=$ORG_ID&metrics=compute_unit_seconds,root_branch_bytes_month,child_branch_bytes_month,instant_restore_bytes_month,public_network_transfer_bytes,private_network_transfer_bytes,extra_branches_month' \
     --header 'Accept: application/json' \
     --header 'Authorization: Bearer $NEON_API_KEY' | jq

For API details, see Retrieve project consumption metrics. For more information, see Querying consumption metrics.

Try it with your AI agent:

Simpler MCP Server setup

You can now configure the Neon MCP Server for all detected AI agents and editors in your workspace with a single command:

npx add-mcp https://mcp.neon.tech/mcp

This adds the MCP config to your editor; restart your editor (or enable the MCP server in settings). When you use the connection, an OAuth window will open in your browser to authorize access. For the full setup (MCP server plus agent skills and VS Code extension), use npx neonctl@latest init instead. It configures the MCP server for Cursor, VS Code, Claude Code, and others using API key authentication.

With OAuth, the MCP server uses your personal Neon account by default. To use organization projects, provide org_id or project_id in your prompt. For API key-based authentication (e.g., remote agents), use:

npx add-mcp https://mcp.neon.tech/mcp --header "Authorization: Bearer $NEON_API_KEY"

For more setup options (including global vs project-level), see Connect MCP clients to Neon and the add-mcp repository.

Postgres extension updates

We've expanded extension support for Postgres 18:

To install these extensions, run:

CREATE EXTENSION pg_graphql;
CREATE EXTENSION pgx_ulid;

For a complete list of Postgres extensions supported by Neon, see Postgres extensions.

Neon Auth SDK simplified

We've released a major update to the server-side Neon Auth SDK (@neondatabase/auth/next/server) for Next.js applications.

Unified entry point

The SDK now uses a single createNeonAuth() function that replaces the previous separate functions (neonAuth(), authApiHandler(), neonAuthMiddleware(), createAuthServer()). Configure authentication once and access all functionality from a single object:

// lib/auth/server.ts
import { createNeonAuth } from '@neondatabase/auth/next/server';

export const auth = createNeonAuth({
  baseUrl: process.env.NEON_AUTH_BASE_URL!,
  cookies: { secret: process.env.NEON_AUTH_COOKIE_SECRET! },
});

// Use everywhere in your app
export const { GET, POST } = auth.handler();           // API routes
export default auth.middleware({ loginUrl: '...' });   // Middleware
const { data: session } = await auth.getSession();     // Server components
await auth.signIn.email({ email, password });          // Server actions

Explicit configuration

Configuration is now explicit rather than implicit. You must pass baseUrl and cookies.secret directly to createNeonAuth() instead of relying on automatic environment variable reading, making dependencies clear and eliminating "magic" behavior.

Session caching

Session data is now automatically cached in a signed cookie, reducing API calls to the Auth Server by 95-99%. Sessions are cached for 5 minutes by default (configurable) and automatically refresh as needed.

Breaking changes

This release includes breaking changes. If you're using Neon Auth and want to upgrade to the latest SDK version, you will need to update your application code. Key changes include replacing separate auth functions with the unified createNeonAuth() API, adding a required NEON_AUTH_COOKIE_SECRET environment variable, and adding dynamic = 'force-dynamic' to server components that use auth methods.

For detailed migration instructions, see the migration guide. For complete API documentation, see the Next.js Server SDK reference.

Getting started

To see the latest SDK in action, check out the demo applications, including a Next.js demo with server components, a React + Vite demo with external UI, and a React demo with the full neon-js SDK. The repository also includes AI coding assistant skills for Cursor and Claude Code with updated setup guides and code examples. See the Neon Auth quickstart guide and Server SDK reference to get started.

Neon Auth is a managed authentication service that branches with your database. See the Neon Auth overview to learn more.

Instagres adds REST API

Instagres now offers a REST API for programmatic database provisioning, making it easy to integrate Postgres into your platform, CI/CD pipelines, testing frameworks, and automation workflows.

The new API enables you to create databases with a single HTTP request:

curl -X POST https://instagres.com/api/v1/database \
  -H 'Content-Type: application/json' \
  -d '{"ref": "your-app-name"}'

The API returns the ID, status, Neon project ID, connection string, claim URL, expiration timestamp, and creation/update timestamps. You can also retrieve database details using GET /api/v1/database/:id. Unclaimed databases have a 100 MB storage limit and expire after 72 hours. Claim your database to a Neon account to remove the expiration and get full Free plan limits.

The get-db CLI also adds a new --logical-replication flag to enable logical replication for real-time sync with tools like ElectricSQL.

Instagres provides instant cloud-hosted Postgres that spins up in seconds. No signup or registration required. See the Instagres documentation for more information.

New NAT gateway IP addresses

We've added new NAT gateway IP addresses in the AWS US East (N. Virginia), US East (Ohio), and US West (Oregon) regions to expand infrastructure capacity. If you have IP allowlists on external systems that Neon connects to, update those allowlists to include the new addresses. Connections may be affected intermittently if traffic routes through non-allowlisted NAT gateways.

See our Regions documentation for the complete list of NAT gateway IPs for all regions.

New VPC endpoint services for Private Networking

We've added new VPC endpoint service addresses for Private Networking in the AWS US East (N. Virginia), US East (Ohio), and US West (Oregon) regions. If you've set up Private Networking in these regions, you can now use the additional endpoint service addresses for enhanced infrastructure capacity and reliability.

For the complete list of VPC endpoint service addresses by region, see our Private Networking guide.