Neon Release Notes

Neon Realtime is in public beta

Neon now has realtime, built by the team behind Electric. Neon Realtime turns any SQL query into a live query. Subscribe to it and the result stays up to date as your data changes. That makes your whole app reactive, with no polling or refetch logic.

It runs directly from Lakebase Postgres, and it works with scale to zero and branching. Out of the box you get type-safe, end-to-end reactivity, SSR support, and instant local writes.

To see what it can do, try one of the live demo apps:

• 1 Million Boxes: the world's largest game of dots and boxes, with every move synced to everyone playing
• Hacker News Realtime: watch Hacker News scores, comment counts, and rankings change as they happen
• Neonbricks: build a brick model together, or ask the AI to build one for you

The source code is available from our Realtime examples page.

Agents can install the Realtime skill with neon skills -s neon-realtime, or use the Neon MCP Server.

To try it yourself, enable Realtime and subscribe to a query:

neon realtime enable
neon realtime subscribe "SELECT * FROM items"

The Realtime quickstart walks through the full setup, or you can start from a template with neon bootstrap my-app --template realtime-react.

During the beta, Realtime is available in AWS us-east-1, us-east-2, eu-central-1, and ap-southeast-1. See the Realtime FAQ for what's in the beta. Give it a try and tell us what you think on Discord.

New models in the AI Gateway

We've added 11 models to the AI Gateway, including new releases from Anthropic, OpenAI, Google, and xAI:

• Anthropic: claude-sonnet-5-5, claude-haiku-5-5
• OpenAI: gpt-6-sol, gpt-6-luna, gpt-6-1-sol
• Google: gemini-3-7-flash, gemini-3-8-flash
• xAI: grok-4-7
• Open-weight models: deepseek-v4-1-flash, glm-5-3 (Zhipu AI), bge-large-en (BAAI embedding model)

If you haven't used the AI Gateway yet, it gives you one endpoint and one Neon credential for models from many providers, so you don't need a separate account or API key for each one. The chat completions endpoint is OpenAI-compatible, so the OpenAI SDK and other standard clients work by changing the base URL. To switch models, change the model field:

const client = new OpenAI({
  apiKey: process.env.NEON_AI_GATEWAY_TOKEN,
  baseURL: `${process.env.NEON_AI_GATEWAY_BASE_URL}/v1`,
});
const response = await client.chat.completions.create({
  model: 'claude-sonnet-5-5',
  messages: [{ role: 'user', content: 'Hello!' }],
});

The AI Gateway is available on the Launch and Scale plans with prepaid credits, which give you access to every model in the catalog. To get started, follow the AI Gateway quickstart. The model catalog lists all supported models.

Sync Stripe data to Neon in real time

Stripe Data Pipeline can now stream your Stripe data straight into Postgres, and Neon is one of the destinations you can pick in the Stripe Dashboard. Stripe backfills your existing data and keeps it in sync as payments, subscriptions, and customers change, so you don't need webhook handlers or Stripe API calls to keep your copy current.

The data lands in the same database as your app, so checking whether a user has an active subscription is a join:

select s.status
from users u
join stripe.subscriptions s on s.customer = u.stripe_customer_id
where u.id = $1
and s.status in ('active', 'trialing');

Real-time sync to Postgres is in public preview, and you can try it for free. To set it up, add a Neon pipeline from the Data Pipeline page in the Stripe Dashboard, choose your Neon project, and pick the tables to sync. To learn more, read Sync Stripe data in real-time to Neon, or see the Stripe docs.

Managed Better Auth upgraded to Better Auth 1.7

Managed Better Auth, Neon's managed authentication service built on the open-source Better Auth library, now runs Better Auth 1.7.6. Your existing apps keep working with no changes. Sessions, cookies, and JWTs are unaffected, and older SDK versions are still supported.

We've also released new SDK versions built on Better Auth 1.7.6: @neondatabase/auth 0.6.0-beta, @neondatabase/neon-js 0.8.0-beta, and @neondatabase/auth-ui 0.4.0-beta.

If you use a Better Auth client of your own with Managed Better Auth, use Better Auth 1.7 or later.

Upgrade note: With the new SDKs, accountInfo, unlinkAccount, getAccessToken, and refreshToken identify a linked account by its local account id, as returned by listAccounts(), instead of the provider's account ID. They no longer accept providerId. If you call these methods directly, update your code:

const { data: accounts } = await authClient.listAccounts();
const google = accounts.find((a) => a.providerId === 'google');
await authClient.accountInfo({ query: { accountId: google.id } });

Known issue: In @neondatabase/auth-ui 0.4.0-beta, the Providers settings card can't unlink an account or show the provider email. A fix is in progress.

Chinese, Japanese, and Korean full-text search in Lakebase Search

Lakebase Search now supports keyword search over Chinese, Japanese, and Korean (CJK) text. These languages don't put spaces between words, so Postgres's built-in full-text search can't split them into search terms. The lakebase_tokenizer extension now does that for you, with ready-to-use tokenizer_chinese, tokenizer_japanese, and tokenizer_korean configurations:

CREATE EXTENSION IF NOT EXISTS lakebase_tokenizer;
SELECT to_tsvector('tokenizer_japanese', '関西国際空港に行きました');
-- '国際':2 '空港':3 '行く':4 '関西':1

The output is a standard tsvector, so it works with GIN indexes and with lakebase_text for BM25 ranking. CJK support ships in lakebase_tokenizer 0.2.0. If you're already using the extension, run ALTER EXTENSION lakebase_tokenizer UPDATE;. Your existing configurations and indexes don't need to change. See CJK text for more examples.

Fixes & improvements

Neon MCP Server

Neon CLI and SDKs

Neon Realtime is in public beta

Neon now has realtime, built by the team behind Electric. Neon Realtime turns any SQL query into a live query. Subscribe to it and the result stays up to date as your data changes. That makes your whole app reactive, with no polling or refetch logic.

It runs directly from Lakebase Postgres, and it works with scale to zero and branching. Out of the box you get type-safe, end-to-end reactivity, SSR support, and instant local writes.

To see what it can do, try one of the live demo apps:

• 1 Million Boxes: the world's largest game of dots and boxes, with every move synced to everyone playing
• Hacker News Realtime: watch Hacker News scores, comment counts, and rankings change as they happen
• Neonbricks: build a brick model together, or ask the AI to build one for you

The source code is available from our Realtime examples page.

Agents can install the Realtime skill with neon skills -s neon-realtime, or use the Neon MCP Server.

To try it yourself, enable Realtime and subscribe to a query:

neon realtime enable
neon realtime subscribe "SELECT * FROM items"

The Realtime quickstart walks through the full setup, or you can start from a template with neon bootstrap my-app --template realtime-react.

During the beta, Realtime is available in AWS us-east-1, us-east-2, eu-central-1, and ap-southeast-1. See the Realtime FAQ for what's in the beta. Give it a try and tell us what you think on Discord.

New models in the AI Gateway

We've added 11 models to the AI Gateway, including new releases from Anthropic, OpenAI, Google, and xAI:

• Anthropic: claude-sonnet-5-5, claude-haiku-5-5
• OpenAI: gpt-6-sol, gpt-6-luna, gpt-6-1-sol
• Google: gemini-3-7-flash, gemini-3-8-flash
• xAI: grok-4-7
• Open-weight models: deepseek-v4-1-flash, glm-5-3 (Zhipu AI), bge-large-en (BAAI embedding model)

If you haven't used the AI Gateway yet, it gives you one endpoint and one Neon credential for models from many providers, so you don't need a separate account or API key for each one. The chat completions endpoint is OpenAI-compatible, so the OpenAI SDK and other standard clients work by changing the base URL. To switch models, change the model field:

const client = new OpenAI({
  apiKey: process.env.NEON_AI_GATEWAY_TOKEN,
  baseURL: `${process.env.NEON_AI_GATEWAY_BASE_URL}/v1`,
});
const response = await client.chat.completions.create({
  model: 'claude-sonnet-5-5',
  messages: [{ role: 'user', content: 'Hello!' }],
});

The AI Gateway is available on the Launch and Scale plans with prepaid credits, which give you access to every model in the catalog. To get started, follow the AI Gateway quickstart. The model catalog lists all supported models.

Sync Stripe data to Neon in real time

Stripe Data Pipeline can now stream your Stripe data straight into Postgres, and Neon is one of the destinations you can pick in the Stripe Dashboard. Stripe backfills your existing data and keeps it in sync as payments, subscriptions, and customers change, so you don't need webhook handlers or Stripe API calls to keep your copy current.

The data lands in the same database as your app, so checking whether a user has an active subscription is a join:

select s.status
from users u
join stripe.subscriptions s on s.customer = u.stripe_customer_id
where u.id = $1
and s.status in ('active', 'trialing');

Real-time sync to Postgres is in public preview, and you can try it for free. To set it up, add a Neon pipeline from the Data Pipeline page in the Stripe Dashboard, choose your Neon project, and pick the tables to sync. To learn more, read Sync Stripe data in real-time to Neon, or see the Stripe docs.

Managed Better Auth upgraded to Better Auth 1.7

Managed Better Auth, Neon's managed authentication service built on the open-source Better Auth library, now runs Better Auth 1.7.6. Your existing apps keep working with no changes. Sessions, cookies, and JWTs are unaffected, and older SDK versions are still supported.

We've also released new SDK versions built on Better Auth 1.7.6: @neondatabase/auth 0.6.0-beta, @neondatabase/neon-js 0.8.0-beta, and @neondatabase/auth-ui 0.4.0-beta.

If you use a Better Auth client of your own with Managed Better Auth, use Better Auth 1.7 or later.

Upgrade note: With the new SDKs, accountInfo, unlinkAccount, getAccessToken, and refreshToken identify a linked account by its local account id, as returned by listAccounts(), instead of the provider's account ID. They no longer accept providerId. If you call these methods directly, update your code:

const { data: accounts } = await authClient.listAccounts();
const google = accounts.find((a) => a.providerId === 'google');
await authClient.accountInfo({ query: { accountId: google.id } });

Known issue: In @neondatabase/auth-ui 0.4.0-beta, the Providers settings card can't unlink an account or show the provider email. A fix is in progress.

Chinese, Japanese, and Korean full-text search in Lakebase Search

Lakebase Search now supports keyword search over Chinese, Japanese, and Korean (CJK) text. These languages don't put spaces between words, so Postgres's built-in full-text search can't split them into search terms. The lakebase_tokenizer extension now does that for you, with ready-to-use tokenizer_chinese, tokenizer_japanese, and tokenizer_korean configurations:

CREATE EXTENSION IF NOT EXISTS lakebase_tokenizer;
SELECT to_tsvector('tokenizer_japanese', '関西国際空港に行きました');
-- '国際':2 '空港':3 '行く':4 '関西':1

The output is a standard tsvector, so it works with GIN indexes and with lakebase_text for BM25 ranking. CJK support ships in lakebase_tokenizer 0.2.0. If you're already using the extension, run ALTER EXTENSION lakebase_tokenizer UPDATE;. Your existing configurations and indexes don't need to change. See CJK text for more examples.

Fixes & improvements

• Neon MCP Server
• Neon CLI and SDKs

All AI Gateway models now available with prepaid credits

You don't need to request access to frontier models in the

Neon AI Gateway

anymore. If you're on a paid plan and have prepaid credits, you can call every model in the catalog, including models from OpenAI, Anthropic, Google, Meta, and Moonshot AI. We also just added Claude Opus 5.5 (claude-opus-5-5). The

Models

page has the full list.

To try it,

buy some prepaid credits

, then follow the

AI Gateway quickstart

to make your first request.

Try Neon's backend features with your coding agent

If you haven't tried

Functions

,

Object Storage

, or the

AI Gateway

yet, your coding agent is a good place to start. First, install Neon's

agent skills

in your project so your agent (Cursor, Claude Code, Codex, or another) knows how these services work:

npx neon@latest skills -y

Then give it something to build. Here are a few ideas:

• Add file uploads to my app with Neon Object Storage, and save each file's metadata in Postgres.
• Create a Neon Function that uses the AI Gateway to summarize support tickets.
• Schedule a Neon Function to run every night and delete files older than 30 days from my Object Storage bucket.

If you'd rather see the bigger picture first,

How a Neon backend fits together

walks through how these services work with Lakebase Postgres and Managed Better Auth.

Send feedback from the CLI and MCP Server

You can now send us feedback straight from your terminal or your coding agent.

Neon MCP Server: Ask your agent to send it with the new send_feedback tool. For example: "send Neon feedback that I'd love a TypeScript example for branching in the docs." See

Neon MCP Server

.

Neon CLI: Run neon feedback --message "". See

neon feedback

.

Either way, we only get your message. No account, project, or connection details come with it.

You can also use the

Feedback

form in the Neon Console, or join the conversation in the

feedback channel

on our Discord.

Fixes & improvements

Here are the smaller updates and fixes we shipped this week.

Console

Neon CLI

Managed Better Auth

New NAT gateway IPs and Private Networking endpoints in eu-central-1

All AI Gateway models now available with prepaid credits

You don't need to request access to frontier models in the Neon AI Gateway anymore. If you're on a paid plan and have prepaid credits, you can call every model in the catalog, including models from OpenAI, Anthropic, Google, Meta, and Moonshot AI. We also just added Claude Opus 5.5 (claude-opus-5-5). The Models page has the full list.

To try it, buy some prepaid credits, then follow the AI Gateway quickstart to make your first request.

Try Neon's backend features with your coding agent

If you haven't tried Functions, Object Storage, or the AI Gateway yet, your coding agent is a good place to start. First, install Neon's agent skills in your project so your agent (Cursor, Claude Code, Codex, or another) knows how these services work:

npx neon@latest skills -y

Then give it something to build. Here are a few ideas:

• Add file uploads to my app with Neon Object Storage, and save each file's metadata in Postgres.
• Create a Neon Function that uses the AI Gateway to summarize support tickets.
• Schedule a Neon Function to run every night and delete files older than 30 days from my Object Storage bucket.

If you'd rather see the bigger picture first, How a Neon backend fits together walks through how these services work with Lakebase Postgres and Managed Better Auth.

Send feedback from the CLI and MCP Server

You can now send us feedback straight from your terminal or your coding agent.

Neon MCP Server: Ask your agent to send it with the new send_feedback tool. For example: "send Neon feedback that I'd love a TypeScript example for branching in the docs." See Neon MCP Server.

Neon CLI: Run neon feedback --message "<your feedback>". See neon feedback.

Either way, we only get your message. No account, project, or connection details come with it.

You can also use the Feedback form in the Neon Console, or join the conversation in the feedback channel on our Discord.

Fixes & improvements

Here are the smaller updates and fixes we shipped this week.

Console

Neon CLI

Managed Better Auth

New NAT gateway IPs and Private Networking endpoints in eu-central-1

More storage on the Free plan

The Neon Free plan now gives you 1 GB of database storage per project, up from 0.5 GB. Existing projects pick up the new limit automatically. It was one of the most requested Free plan changes.

The extra storage builds on an already-generous plan: you get up to 100 projects on the Free plan, each with its own compute, storage, and egress. Spin up a new project for every app idea, experiment, or client demo, without paying for each project separately.

Read more in the blog post:

Neon gives you 100 projects for free, with 1 GB of Postgres storage each, or see Neon plans for the full breakdown.

Embedding models in the Neon AI Gateway

The Neon AI Gateway now supports embedding models, so you can turn text into vectors through the same OpenAI-compatible endpoint and Neon credential you already use for chat. Call /v1/embeddings with either model:

import OpenAI from 'openai';
const client = new OpenAI({
  apiKey: process.env.NEON_AI_GATEWAY_TOKEN,
  baseURL: `${process.env.NEON_AI_GATEWAY_BASE_URL}/v1`,
});
const { data } = await client.embeddings.create({
  model: 'qwen3-embedding-0-6b',
  // or 'gte-large-en'
  input: 'The quick brown fox jumps over the lazy dog.',
  encoding_format: 'float',
});
// data[0].embedding is a 1024-dimension vector

Both models (qwen3-embedding-0-6b and gte-large-en) return 1024-dimension vectors and are billed on input tokens only. Store and search the vectors on the same branch as your data with Lakebase Search or pgvector. For the full reference, including the dimensions parameter and response format, see the Embeddings guide.

Effect bindings for the Neon SDK

Effect is a TypeScript library for building resilient apps, with typed errors, retries, timeouts, and cancellation built in. The new @neon/effect package wraps the Neon Management SDK in Effect v4, so managing your Neon projects fits that model: every API call returns an Effect, paginated lists return a Stream, and SDK errors are tagged for Effect.catchTag. Interrupting a fiber cancels the in-flight request and readiness polling, so retries, timeouts, and cancellation compose with the rest of your program.

import { Neon, layerConfig } from '@neon/effect';
import { Effect, Stream } from 'effect';
const program = Effect.gen(function* () {
  const neon = yield* Neon;
  return yield* neon.projects.list().pipe(Stream.take(10), Stream.runCollect);
});
Effect.runPromise(program.pipe(Effect.provide(layerConfig)));

Install @neon/effect alongside effect v4. Read more in the blog post: Announcing @neon/effect, or see the Effect bindings reference.

Fixes & improvements

• Postgres extension updates
• AI Gateway enhancements
• Neon Functions improvements
• Serverless driver
• Neon Console Tables page updates
• Neon MCP server updates
• Neon CLI and agent skill updates

More storage on the Free plan

The Neon Free plan now gives you 1 GB of database storage per project, up from 0.5 GB. Existing projects pick up the new limit automatically. It was one of the most requested Free plan changes.

The extra storage builds on an already-generous plan: you get up to 100 projects on the Free plan, each with its own compute, storage, and egress. Spin up a new project for every app idea, experiment, or client demo, without paying for each project separately.

Read more in the blog post: Neon gives you 100 projects for free, with 1 GB of Postgres storage each, or see Neon plans for the full breakdown.

Embedding models in the Neon AI Gateway

The Neon AI Gateway now supports embedding models, so you can turn text into vectors through the same OpenAI-compatible endpoint and Neon credential you already use for chat. Call /v1/embeddings with either model:

import OpenAI from 'openai';
const client = new OpenAI({
  apiKey: process.env.NEON_AI_GATEWAY_TOKEN,
  baseURL: `${process.env.NEON_AI_GATEWAY_BASE_URL}/v1`,
});
const { data } = await client.embeddings.create({
  model: 'qwen3-embedding-0-6b', // or 'gte-large-en'
  input: 'The quick brown fox jumps over the lazy dog.',
  encoding_format: 'float',
});
// data[0].embedding is a 1024-dimension vector

Both models (qwen3-embedding-0-6b and gte-large-en) return 1024-dimension vectors and are billed on input tokens only. Store and search the vectors on the same branch as your data with Lakebase Search or pgvector. For the full reference, including the dimensions parameter and response format, see the Embeddings guide.

Effect bindings for the Neon SDK

Effect is a TypeScript library for building resilient apps, with typed errors, retries, timeouts, and cancellation built in. The new @neon/effect package wraps the Neon Management SDK in Effect v4, so managing your Neon projects fits that model: every API call returns an Effect, paginated lists return a Stream, and SDK errors are tagged for Effect.catchTag. Interrupting a fiber cancels the in-flight request and readiness polling, so retries, timeouts, and cancellation compose with the rest of your program.

import { Neon, layerConfig } from '@neon/effect';
import { Effect, Stream } from 'effect';

const program = Effect.gen(function* () {
  const neon = yield* Neon;
  return yield* neon.projects.list().pipe(Stream.take(10), Stream.runCollect);
});

Effect.runPromise(program.pipe(Effect.provide(layerConfig)));

Install @neon/effect alongside effect v4. Read more in the blog post: Announcing @neon/effect, or see the Effect bindings reference.

Fixes & improvements

• Postgres extension updates
• AI Gateway enhancements
• Neon Functions improvements
• Serverless driver
• Neon Console Tables page updates
• Neon MCP server updates
• Neon CLI and agent skill updates

Options to the neon(...) HTTP-client creation function, and to the transaction and query functions, now support individual connection parameters: user (or username), password, host (or hostname), port, database and connectionString.

These can be strings (or number, for port) or sync or async functions to be resolved on every query. They can supplement, override parts of, or fully replace the first-argument connection string.

So this still works:

• const sql = neon(DATABASE_URL)

But, for example, these now work too:

• const sql = neon(DATABASE_URL_NO_PASSWORD, { password: () => getPasswordAsync() })
• const sql = neon({ user, password, host, database })
• const sql = neon({ connectionString: () => getConnectionString() })

Client and Pool now support a sync or async function as the password connection parameter when pipelineConnect is enabled (this previously caused an error).

And parseIntoClientConfig from pg-connection-string is exported at top-level to make it easy to use a password function alongside a connection string:

import { Pool, parseIntoClientConfig } from '@neondatabase/serverless';

const pool = new Pool({
  ...parseIntoClientConfig(process.env.DATABASE_URL),
  password: () => getDatabaseAuthToken('argument'),
});

A few other minor fixes.

The package is no longer published to JSR.

Please update any jsr references to npm:@neondatabase/serverless.

Reset from parent now rolls back a branch's Object Storage along with its Postgres data.

When you reset a branch, its buckets and objects return to the parent's current state, so any object storage uploads, overwrites, and deletes you made on the branch are also reverted.

You can reset a branch from the Neon Console, CLI, or API. See Reset from parent for the steps.

Resetting Object Storage is available now in AWS US East (Ohio) (aws-us-east-2) and AWS Europe (Frankfurt) (aws-eu-central-1), with support for additional regions rolling out soon.

The Lakebase Search extensions are updated: lakebase_vector to 1.1.1 and lakebase_text to 0.1.3, with performance improvements and fixes. If you have them installed, upgrade when convenient with ALTER EXTENSION <name> UPDATE; see Upgrade the extension and indexes.

NOTE

Building search with an AI coding assistant? The neon-postgres agent skill now includes Lakebase Search guidance for vector, full-text, and hybrid queries. Install it with:

neon skills -s neon-postgres

Automating workloads with Neon Functions

Last week we announced that the Neon backend is generally available, with Functions running next to your database. This week, a new guide shows how to use Neon Functions to automate workloads with Function Triggers: run a Function on a UTC cron schedule or when a file lands in an Object Storage bucket.

Learn how to declare a function in code using a neon.ts file and apply it with neon deploy.

// neon.ts
functions: {
  reports: { name: 'Reports', source: './functions/reports.ts' },
},
triggers: {
  nightly: {
    type: 'schedule',
    function: 'reports',
    cron: '0 6 * * *',
  },
},

Realtime is coming to Neon

Realtime sync is coming to Neon. The team behind Electric, the Postgres sync engine, joined Neon at Databricks and is building native realtime and end-to-end reactivity into the platform, designed to work with branching and scale to zero. Read our Q&A with Electric co-founder James Arthur.

Reset a branch's Object Storage from its parent

Reset from parent now rolls back a branch's Object Storage along with its Postgres data. When you reset a branch, its buckets and objects return to the parent's current state, so any object storage uploads, overwrites, and deletes you made on the branch are also reverted.

You can reset a branch from the Neon Console, CLI, or API. See Reset from parent for the steps.

Resetting Object Storage is available now in AWS US East (Ohio) (aws-us-east-2) and AWS Europe (Frankfurt) (aws-eu-central-1), with support for additional regions rolling out soon.

Lakebase Search extensions updated

The Lakebase Search extensions are updated: lakebase_vector to 1.1.1 and lakebase_text to 0.1.3, with performance improvements and fixes. If you have them installed, upgrade when convenient with ALTER EXTENSION UPDATE; see Upgrade the extension and indexes.

NOTE

Building search with an AI coding assistant? The neon-postgres agent skill now includes Lakebase Search guidance for vector, full-text, and hybrid queries. Install it with:
neon skills -s neon-postgres

Automating workloads with Neon Functions

Last week we announced that the Neon backend is generally available, with Functions running next to your database. This week, a new guide shows how to use Neon Functions to automate workloads with Function Triggers: run a Function on a UTC cron schedule or when a file lands in an Object Storage bucket.

Learn how to declare a function in code using a neon.ts file and apply it with neon deploy.

// neon.ts
functions: {
reports: { name: 'Reports', source: './functions/reports.ts' },
},
triggers: {
nightly: {
type: 'schedule',
function: 'reports',
cron: '0 6 * * *',
},
},

Realtime is coming to Neon

Realtime sync is coming to Neon. The team behind Electric, the Postgres sync engine, joined Neon at Databricks and is building native realtime and end-to-end reactivity into the platform, designed to work with branching and scale to zero. Read our Q&A with Electric co-founder James Arthur.

The Neon backend is generally available

Neon started as serverless Postgres. Now you get the whole backend, and all of it branches. Every branch is a full copy, so a PR, preview, or agent session gets your entire backend, not just the database. Evolve your schema, storage, auth, and functions together on a preview branch, isolated from production.

▶️ Watch the walkthrough to see how it all fits together.

• Lakebase Postgres 100% Postgres on the Lakebase architecture: autoscaling, scale to zero, and a branch for every PR, preview, or agent session.
• Object Storage S3-compatible object storage that branches with your data, so files and rows stay in sync across branches.
• Auth Managed Better Auth: users and sessions live in your neon_auth schema, queryable with SQL and protected by RLS.
• Functions Long-running Node.js compute next to your database, now with triggers and custom domains.
• AI Gateway One credential for frontier and open-weight models at provider prices with no markup. Buy prepaid credits to use it.
• Data API An instant REST endpoint over your Postgres tables, no backend server required.

How to get started

Building with a coding agent? neon skills installs the Neon agent skills into your assistant (Cursor, Claude Code, Codex, and more), so it knows how to provision and wire the whole backend.

However you build, your whole backend is declared in one neon.ts file and applied with neon deploy, whether your agent writes it or you do:

New with the GA release

Function Triggers

Turn an upload into a pipeline or run a job on a schedule, without standing up a queue or a cron server. A Function Trigger invokes a deployed Function on a schedule or an event, and fires even when the compute is scaled to zero. Two trigger types are available now, with more on the way:

• Schedule: a UTC cron expression, for recurring work like a nightly rollup or a health check.
• Object created: fires when a file lands in an Object Storage bucket (optionally under a key prefix), so an upload can kick off processing like generating a thumbnail.

Declare a trigger in neon.ts, or use the Console, CLI, or API:

Custom domains for Functions

Ship your public API or webhook on a domain you own, like api.example.com, instead of the native Neon URL. Point a CNAME at the target Neon gives you, and Neon provisions the TLS certificate automatically. Declare it in neon.ts:

Domains are branch-scoped, so a preview branch can serve from its own hostname while production keeps yours. See Custom domains.

Prepaid credits for the AI Gateway

You can now buy prepaid credits for the AI Gateway from the Console. Credits draw down as you send inference requests, at provider list prices with no markup. See AI Gateway prepaid credits.

More regions

Object Storage, Functions, and AI Gateway now run in four AWS regions: US East (Ohio), US East (N. Virginia), Europe (Frankfurt), and Asia Pacific (Singapore), with more coming.

Turn on the backend when you create a project

Adopting the backend now starts at project creation. You can now enable Object Storage, Functions, AI Gateway, and Auth alongside Postgres when you first create your project, so it comes up with your whole stack already wired. You will only see services that are supported in your selected region.

A redesigned project settings page

The project Settings page has a cleaner, tabbed layout (General, Postgres, Networking, Credentials, Auth, Functions, HIPAA compliance, and Sharing) that groups related controls and labels project-wide settings as Applies to all branches. Setting your default branch on the General tab is now project-scoped: use Change default branch to pick any branch.

Organizations and account settings moved to the profile menu

Switching and creating organizations, organization management, account settings, theme, Neon status, and help now live in one profile menu in the top-right corner of the Console.

Fixes & improvements

• SDK
• neon.ts config
• CLI
• New NAT gateway IPs and Private Networking endpoints in sa-east-1

Neon started as serverless Postgres. Now you get the whole backend, and all of it branches. Every branch is a full copy, so a PR, preview, or agent session gets your entire backend, not just the database. Evolve your schema, storage, auth, and functions together on a preview branch, isolated from production.

▶️ Watch the walkthrough to see how it all fits together.

Lakebase Postgres

100% Postgres on the Lakebase architecture: autoscaling, scale to zero, and a branch for every PR, preview, or agent session.

Object Storage

S3-compatible object storage that branches with your data, so files and rows stay in sync across branches.

Auth

Managed Better Auth: users and sessions live in your neon_auth schema, queryable with SQL and protected by RLS.

Functions

Long-running Node.js compute next to your database, now with triggers and custom domains.

AI Gateway

One credential for frontier and open-weight models at provider prices with no markup. Buy prepaid credits to use it.

Data API

An instant REST endpoint over your Postgres tables, no backend server required.

How to get started

Building with a coding agent? neon skills installs the Neon agent skills into your assistant (Cursor, Claude Code, Codex, and more), so it knows how to provision and wire the whole backend.

However you build, your whole backend is declared in one neon.ts file and applied with neon deploy, whether your agent writes it or you do:

// neon.ts
import { defineConfig } from '@neon/config/v1';

export default defineConfig({
  auth: true, // managed authentication
  buckets: { uploads: {} }, // S3-compatible object storage
  functions: { api: { source: './functions/api.ts' } }, // serverless compute
  aiGateway: true, // one API for many LLMs
});

Get started with How a Neon backend fits together or the full-stack quickstart, or scaffold a working example with neon bootstrap. Read the launch blog post for the full story.

New with the GA release

Function Triggers

Turn an upload into a pipeline or run a job on a schedule, without standing up a queue or a cron server. A Function Trigger invokes a deployed Function on a schedule or an event, and fires even when the compute is scaled to zero. Two trigger types are available now, with more on the way:

• Schedule: a UTC cron expression, for recurring work like a nightly rollup or a health check.
• Object created: fires when a file lands in an Object Storage bucket (optionally under a key prefix), so an upload can kick off processing like generating a thumbnail.

Declare a trigger in neon.ts, or use the Console, CLI, or API:

// neon.ts
functions: {
  reports: {
    source: './functions/reports.ts',
    triggers: [{ type: 'schedule', name: 'nightly', cron: '0 6 * * *' }],
  },
},

See Function Triggers.

Custom domains for Functions

Ship your public API or webhook on a domain you own, like api.example.com, instead of the native Neon URL. Point a CNAME at the target Neon gives you, and Neon provisions the TLS certificate automatically. Declare it in neon.ts:

// neon.ts
functions: {
  api: {
    source: './functions/api.ts',
    customDomains: ['api.example.com'],
  },
},

Domains are branch-scoped, so a preview branch can serve from its own hostname while production keeps yours. See Custom domains.

Prepaid credits for the AI Gateway

You can now buy prepaid credits for the AI Gateway from the Console. Credits draw down as you send inference requests, at provider list prices with no markup. See AI Gateway prepaid credits.

More regions

Object Storage, Functions, and AI Gateway now run in four AWS regions: US East (Ohio), US East (N. Virginia), Europe (Frankfurt), and Asia Pacific (Singapore), with more coming.

Turn on the backend when you create a project

Adopting the backend now starts at project creation. You can now enable Object Storage, Functions, AI Gateway, and Auth alongside Postgres when you first create your project, so it comes up with your whole stack already wired. You will only see services that are supported in your selected region.

A redesigned project settings page

The project Settings page has a cleaner, tabbed layout (General, Postgres, Networking, Credentials, Auth, Functions, HIPAA compliance, and Sharing) that groups related controls and labels project-wide settings as Applies to all branches. Setting your default branch on the General tab is now project-scoped: use Change default branch to pick any branch.

Organizations and account settings moved to the profile menu

Switching and creating organizations, organization management, account settings, theme, Neon status, and help now live in one profile menu in the top-right corner of the Console.

Fixes & improvements

• SDK
• neon.ts config
• CLI
• New NAT gateway IPs and Private Networking endpoints in sa-east-1

Claimable Neon: provisioned by agents, claimed by humans

Claimable Neon is for developers building with agents. When your agent is mid-build, the signup, email verification, or API key it needs to provision a database brings everything to a halt. Claimable Neon solves this problem. It lets your agent provision a temporary project with scoped credentials, without an account or payment details, and keep building. If you like the result, you can sign in later and claim the project.

Agent: provision a claimable project (writes DATABASE_URL to .env)

neon claim create --env-pull

Human: sign in and take ownership

neon claim accept

Today you can provision Lakebase Postgres, the Data API, and Managed Better Auth this way, with support for Object Storage, Functions, and AI Gateway coming soon.

For how it works, see the Claimable Neon docs and the launch blog post.

Give your agent Neon tools with @neon/tools

If you're building an AI agent or app that manages Neon itself, such as one that spins up a project or a branch, @neon/tools gives it typed, function-calling tools instead of hand-written API wrappers. It turns the @neon/sdk client into tools you hand to a model. Pick the Neon operations you want to expose, and they stay in sync with Neon's API. It also ships adapters for MCP, Mastra, and Eve.

import { createNeonTools } from '@neon/tools';
const tools = createNeonTools({
  apiKey: process.env.NEON_API_KEY,
  tools: [
    'projects.list',
    'projects.createAndConnect',
    'branches.createAndConnect'
  ],
});
// createAndConnect provisions a project, attaches compute, and returns a connection string
const created = await tools['projects.createAndConnect'].execute({
  name: 'agent-project',
  region_id: 'aws-us-east-1',
});

The same @neon/tools catalog now powers the hosted Neon MCP server for coding assistants like Cursor and Claude, adding over 70 new tools. Read more in Give your agent Neon tools.

Faster queries on large computes

Lakebase Postgres now puts much more of a large compute's memory to work as cache. On fixed-size computes of 18 CU and above, shared buffers expand to use up to 75% of available memory, backed by huge pages, so more of your working set stays resident and fewer reads fall through to storage. Cache-heavy workloads can see up to 2x throughput, with no configuration change or restart required.

On an affected compute you can confirm the larger cache with SHOW shared_buffers and SHOW huge_pages.

For the engineering deep dive, see Improving Lakebase Postgres compute cache on Neon, Part 1.

Fixes & improvements

Scheduled snapshots

Claimable Neon: provisioned by agents, claimed by humans

Claimable Neon is for developers building with agents. When your agent is mid-build, the signup, email verification, or API key it needs to provision a database brings everything to a halt. Claimable Neon solves this problem. It lets your agent provision a temporary project with scoped credentials, without an account or payment details, and keep building. If you like the result, you can sign in later and claim the project.

Agent: provision a claimable project (writes DATABASE_URL to .env)

neon claim create --env-pull

Human: sign in and take ownership

neon claim accept

Today you can provision Lakebase Postgres, the Data API, and Managed Better Auth this way, with support for Object Storage, Functions, and AI Gateway coming soon.

For how it works, see the Claimable Neon docs and the launch blog post.

Give your agent Neon tools with @neon/tools

Give your agent Neon tools with @neon/tools

If you're building an AI agent or app that manages Neon itself, such as one that spins up a project or a branch, @neon/tools gives it typed, function-calling tools instead of hand-written API wrappers. It turns the @neon/sdk client into tools you hand to a model. Pick the Neon operations you want to expose, and they stay in sync with Neon's API. It also ships adapters for MCP, Mastra, and Eve.

import { createNeonTools } from '@neon/tools';
const tools = createNeonTools({
  apiKey: process.env.NEON_API_KEY,
  tools: [
    'projects.list',
    'projects.createAndConnect',
    'branches.createAndConnect'
  ],
});

// createAndConnect provisions a project, attaches compute, and returns a connection string
const created = await tools['projects.createAndConnect'].execute({
  name: 'agent-project',
  region_id: 'aws-us-east-1',
});

The same @neon/tools catalog now powers the hosted Neon MCP server for coding assistants like Cursor and Claude, adding over 70 new tools. Read more in Give your agent Neon tools.

Faster queries on large computes

Lakebase Postgres now puts much more of a large compute's memory to work as cache. On fixed-size computes of 18 CU and above, shared buffers expand to use up to 75% of available memory, backed by huge pages, so more of your working set stays resident and fewer reads fall through to storage. Cache-heavy workloads can see up to 2x throughput, with no configuration change or restart required.

On an affected compute you can confirm the larger cache with SHOW shared_buffers and SHOW huge_pages.

For the engineering deep dive, see Improving Lakebase Postgres compute cache on Neon, Part 1.

Fixes & improvements

Scheduled snapshots

Neon backend in Europe, plus new AI Gateway models

The Neon backend is now available in AWS Europe (Frankfurt) (aws-eu-central-1), alongside AWS US East (Ohio) (aws-us-east-2). If you build in Europe, you can now run these services closer to your users and data.

The AI Gateway catalog also gains new models, callable through the same API and key as the rest of the catalog:

• Fable 5.1
• GLM 5.3 Flash
• Grok 4.6
• OpenAI GPT-6 Astra (landing soon)

Try the Neon backend, free during the beta

Object Storage, Functions, and AI Gateway are in beta in AWS US East (Ohio) and AWS Europe (Frankfurt), and branch with your Postgres database, so your whole stack forks together.

Neon is now available in Grok Bot

Neon is now available as a plugin in Grok Bot. Connect your Neon account once, then ask a Grok bot to work with your database:

• Create a project and get a connection string for your app
• Spin up an isolated branch before testing a migration
• List projects and spot databases you no longer need
• Run SQL against your database
• Set up routines that summarize database usage

The plugin pairs Neon agent skills with the Neon MCP server, so Grok Bot has both the context and the authenticated tools to act on your account.

To install, open the Neon plugin in Grok Bot and authorize Neon in the browser. You can also open Grok Bot's Plugins directory and search for Neon, or ask Grok to connect the Neon plugin. Read more in the blog post.

Neon CLI updates

Two updates to the Neon CLI for setting up and connecting your projects:

• A leaner neon init: it now does one focused job based on where you run it: in an empty directory it scaffolds a template and installs agent tooling (like neon bootstrap); in an existing app it installs a plugin (or skills and MCP), links your project, and writes a neon.ts config. Run it with npx, and pass -y and --agent for unattended runs:
  npx neon@latest init
• neon env pull writes Neon Function invocation URLs: each function's public URL is written as NEON_FUNCTION__BASE_URL, so one function can reach another from the environment. See Environment variables.

The Neon backend is now available in AWS Europe (Frankfurt) (aws-eu-central-1), alongside AWS US East (Ohio) (aws-us-east-2)

If you build in Europe, you can now run these services closer to your users and data.

The AI Gateway catalog also gains new models, callable through the same API and key as the rest of the catalog :

• Fable 5.1
• GLM 5.3 Flash
• Grok 4.6
• OpenAI GPT-6 Astra (landing soon)

Try the Neon backend, free during the beta

Object Storage, Functions, and AI Gateway are in beta in AWS US East (Ohio) and AWS Europe (Frankfurt), and branch with your Postgres database, so your whole stack forks together.

Neon is now available in Grok Bot

Neon is now available as a plugin in Grok Bot. Connect your Neon account once, then ask a Grok bot to work with your database:

• Create a project and get a connection string for your app
• Spin up an isolated branch before testing a migration
• List projects and spot databases you no longer need
• Run SQL against your database
• Set up routines that summarize database usage

The plugin pairs Neon agent skills with the Neon MCP server, so Grok Bot has both the context and the authenticated tools to act on your account.

To install, open the Neon plugin in Grok Bot and authorize Neon in the browser. You can also open Grok Bot's Plugins directory and search for Neon, or ask Grok to connect the Neon plugin. Read more in the blog post.

Neon CLI updates

Two updates to the Neon CLI for setting up and connecting your projects:

• A leaner neon init: it now does one focused job based on where you run it: in an empty directory it scaffolds a template and installs agent tooling (like neon bootstrap); in an existing app it installs a plugin (or skills and MCP), links your project, and writes a neon.ts config. Run it with npx, and pass -y and --agent for unattended runs:

npx neon@latest init

• neon env pull writes Neon Function invocation URLs: each function's public URL is written to your local .env as NEON_FUNCTION__BASE_URL, so your local app or frontend can call your deployed functions. See Environment variables.

neon env pull