Modern GPS datasets are notoriously noisy: satellites drift, buildings scatter signals, and consumer devices introduce frequent errors. When working with millions of position samples from vehicles, smartphones, or IoT devices, this noise makes analysis unreliable. Routes jump, tracks zigzag, and out...

Postgres 18 just shipped, and as always, it’s available in preview on Neon right away. You can try it right now on Neon’s Free plan, no credit card or setup required. New Features in Postgres 18 Postgres 18 comes with a ton of other exciting updates around performance, flexibility, and developer exp...

Elephantshark helps you monitor, understand and troubleshoot Postgres network traffic: that’s Postgres servers, clients, drivers and ORMs talking to Postgres servers, proxies and poolers. Elephantshark sits between the two parties in a Postgres-protocol exchange, forwarding messages in both directio...

You can now create Neon snapshots via API. This new capability isn’t just useful for backups or disaster recovery, but also serves as a powerful building block for one of the most requested features in agentic platforms: versioning (or checkpoints). Neon’s snapshots, built on our copy-on-write branc...

You’ve heard this many times before – in order to keep your Postgres database working smoothly, you need to have proper index planning. Too few indexes, and your query performance suffers. Misestimated row counts can also trick the planner into poor choices, and if you’re not using prepared statemen...

“We recently shipped a big update and we probably couldn’t have done it without Neon branching. It lets us test big migrations safely with production data, and if something breaks, we just delete the branch and start fresh. That makes us a lot faster and more confident when shipping, even as a small...

OpenAI just launched support for MCP servers in ChatGPT, letting you connect external tools – including Neon! Via Neon’s MCP server, you can bring your Postgres projects into ChatGPT and ask questions about resource usage, database activity, branches, and more. ChatGPT Now Supports MCP Servers MCP (...

You can mock database calls all day long, but when your code hits production, the real database doesn’t lie. Unique constraints fail, transactions don’t roll back as they should, and subtle behaviors that mocks can’t reproduce, causing bad code to go into production. Running integration tests agains...

Most database providers follow the classic SaaS playbook: bundle features into tiered plans with progressively higher monthly fees. This looks neat on a pricing page. Almost nobody questions this model, but it creates a quagmire inside product teams. We know it from experience, and if you’ve worked ...

Note: This blog post was updated on October 31, 2025 to reflect changes in the Free Plan limits. Neon’s Free Plan is packed with everything you need to start building with Postgres – we’ve seen developers ship amazing things with it. If you’re new to Neon, let us share some tips to make sure you […]

Using AI coding assistants like Cursor to build your full-stack apps makes for a powerful workflow, but they don’t always know the specifics of every tool – including Neon – and can do a little help. To give your editor Neon-specific context and help it generate better code from the start, we’ve cre...

Postgres is one of the most versatile databases out there. You can store vectors, declare row-level security rules, run background jobs, and more. Its simplicity and extensibility make it a workhorse for everything from side projects to large-scale systems. But that flexibility comes with a trade-of...

“I was surprised by how fast Neon was. It was faster than my self-hosted setup and Prisma Postgres. This plus the convenience of the Free Plan makes it a no-brainer for building your projects.” (Issa Nassar, founder of Databuddy) Databuddy is a new privacy-first web and product analytics platform th...

LLMs are fantastic at code generation. They can produce hundreds of lines of code, entire functions and components, that compile perfectly, implement complex algorithms elegantly, and even follow best practices for style and structure. They suck at full-blown software generation, though. Production-...

Postgres is loved not just for what it is, but for what it can become: its extension system lets you transform your database into a full-text search engine, a geospatial powerhouse, or a vector store for embeddings – all without switching tools.  With many thousands of active databases running on ou...

In Neon, we’re always trying to strike a balance between out-of-the-box configurations that work for most people and the flexibility to accommodate a wide range of use cases. By tweaking features like autoscaling, branching, and scale to zero, you can optimize Neon for many different scenarios – fro...

In the Stack Overflow 2025 Developer Survey, “84% of respondents are using or planning to use AI tools in their development process.” They don’t break this number down by developer language, but if they did, they would likely find a wide disparity between different languages. While TS/JS and Python ...

We’ve just rolled out a major update to Neon’s pricing. Starting today, our paid plans are fully usage-based. That means you only pay for what you use, with no quotas to manage, no overages to worry about, and no need to guess how much capacity you’ll need upfront. We’re only enforcing a $5/month mi...

Branching is one of the most powerful features in Neon. It lets teams spin up isolated environments instantly and automate workflows through the API, and as a result, developers and agents create many branches at scale – hundreds or even thousands of them. To make managing short-lived branches even ...

Neptune is an AI-powered concierge that helps couples navigate high-stakes financial decisions, starting with prenups. Neptune guides couples through an emotionally intelligent process that replaces the guesswork and complexity of prenups (because most of us are not lawyers) and helps couples identi...

App.build now supports open weights LLMs via Ollama, LMStudio, and OpenRouter – enabling you to generate complete applications end-to-end without cloud API dependency or associated costs. Why Run App.build Locally? Zero API costs Cloud LLM APIs can become expensive fast during extended coding sessio...

The former Create.xyz has now relaunched as Anything, a new platform agent that builds and ships full-stack applications from a single prompt with support for mobile and web, production-ready designs, and built-in infrastructure, with Neon powering the backend with serverless Postgres, built-in auth...

“The entire memory layer of our agent, all its session data, is stored in Neon. It’s the backend for everything the agent remembers.” (Ran Sheinberg, co-founder and CPO of xpander.ai) xpander is a new platform built for helping developers build and run production-ready AI agents. Today, they’re laun...

Multiplayer UIs aren’t new. Tools like Figma and Google Docs have brought collaborative, real-time experiences to millions of users. But for most developers, building robust and scalable real-time applications has always been out of reach. Real-time is hard. The benefits are obvious: no stale data, ...

Dyad is a local-first, open-source AI coding assistant that helps you build and iterate on web apps from your own machine. You can chat with Dyad to scaffold code, review changes, or debug issues all inside your editor. Unlike most cloud-based tools, Dyad runs locally, supports any LLM (including GP...

More and more software will be created by agents, not just code snippets but full-stack apps with interfaces, backends, persistence, auth. This means a big change in the shape of the devtools and software infrastructure supporting these apps. Many systems are now created at runtime, per app, per req...

We’re very excited to announce that app.build is launching a web interface that builds and deploys React, Laravel or FastAPI applications with zero configuration required! Just visit the website, type your prompt, and we’ll get your app built and deployed: This makes the experience of trying app.bui...

At Laracon 2025, we just demoed a new way to build Laravel apps: by writing a prompt! app.build is an open-source reference architecture for agent codegen, able to generate full-stack web apps from natural language. It handles everything from scaffolding the project to writing tests, deploying to th...

It’s not only hype: developers are working with AI every day – both using new AI-powered tools and building their own. At Neon, we’re seeing Postgres widely used in three AI workflows:  Vector search with pgvector Many AI applications rely on vector embeddings for semantic search, RAG pipelines and ...

Every now and then, people ask me:  “I am new to agentic development, I’m building something, but I feel like I’m missing some tribal knowledge. Help me catch up!” I’m tempted to suggest some serious stuff like multiweek courses (e.g. by HuggingFace or Berkeley), but not everyone is interested in th...