RootbaseRootbase
← DirectorioEntrar

Database / Backend-as-a-Service

Neon

The backend for apps and agents, built to scale on Lakebase Postgres.

neon.tech/ ↗
98
AI Readiness

Neon is the backend platform for apps and agents, built around Lakebase Postgres (a serverless, fully managed Postgres database). Neon became part of the Databricks Platform in May 2025. It was founded by Postgres committers and serves 15 million Postgres databases started daily.

¿Es tu empresa?

Reclámala verificando tu dominio y controla cómo te ven los agentes.

Producto

Neon provides a complete serverless backend suite: Lakebase Postgres (serverless, autoscaling, branching Postgres), Managed Better Auth (authentication stored in Postgres), Neon Functions (long-running serverless compute next to the database), Object Storage (S3-compatible storage that branches with the database), and an AI Gateway (one API for frontier and open-source LLMs). Every service is agent-ready, branchable, and serverless.

Problema que resuelve: Eliminates the need to manage database infrastructure, provision servers, or maintain database clusters. Enables instant branching for development/testing, autoscaling to avoid over-provisioning, and a unified backend so developers don't need separate services for auth, storage, and AI model access.

Para: Application developers building web apps, AI/agent developers building full-stack AI agents, Startups needing scalable Postgres without infrastructure overhead, Platforms offering Postgres to their users, Teams needing database-per-tenant architectures, Developers migrating from RDS, Supabase, PlanetScale, Heroku, and other Postgres providers

Casos de uso

Features (12)

Lakebase Postgres (Serverless Postgres)◦

Fully managed Postgres with storage-compute separation, autoscaling, scale-to-zero, branching, instant restore, and read replicas. Compatible with any Postgres driver, ORM, or framework.

Autoscaling◦

Compute (CPU and memory) scales up and down automatically based on workload demand, including on the Free plan. Each CU allocates approximately 1 vCPU and 4 GB of RAM.

Database Branching◦

Create instant, isolated, copy-on-write copies of any database using git-like branching. Branches can be ephemeral and delete themselves automatically after work is complete.

Managed Better Auth◦

Managed authentication service built into Lakebase Postgres. Stores users, sessions, and OAuth configuration in the neon_auth schema in Postgres. Supports email/password, OAuth (Google), magic links, OTP, and more. Auth state branches with the database.

Neon Functions◦

Long-running serverless compute (Node.js/TypeScript) deployed onto a Neon branch, running in the same region as the database. Supports REST APIs, AI agents, WebSocket/SSE servers, MCP servers, and webhook handlers. No short execution timeout like lambda-style functions.

Object Storage◦

S3-compatible object storage built into the Neon backend. Each branch gets its own isolated storage namespace. Works with AWS S3 SDKs, boto3, AWS CLI, and other S3-compatible tools. Supports private and public_read bucket access modes.

AI Gateway◦

LLM inference layer built into the Neon backend. Provides one API and one credential for models from OpenAI (GPT), Google (Gemini), Meta, Databricks, and Alibaba (Qwen). Compatible with the OpenAI SDK and google-genai SDK with just a URL change. Supports streaming via server-sent events.

Instant Restore (Point-in-Time Recovery)◦

Restore the database to any point in time within the configured history window (up to 30 days on Scale plan) without flat monthly fees. Snapshots are also supported.

Lakebase Architecture (Storage-Compute Separation)◦

Postgres compute and storage are decoupled into independent layers communicating via WAL streams. Safekeepers replicate WAL via Paxos quorum; the pageserver materializes pages on demand; object storage holds immutable long-term history.

Lakebase Search◦

Vector, keyword, and hybrid search built directly into the database via pgvector and pg_search extensions, so agent retrieval and RAG require no separate vector store.

Connection Pooling◦

Built-in connection pooling via pgBouncer, supporting up to 10,000 concurrent connections. Available on all plans by default.

Data Anonymization for Branches◦

Mask sensitive data with realistic fake values when creating branches, enabling safe testing and sharing of datasets.

Pricing

Free

◦

0/month

  • · Lakebase Postgres
  • · Managed Better Auth
  • · Object Storage (beta, no charges during beta)
  • · Functions (beta, no charges during beta)
  • · Autoscaling
  • · Unlimited team members

Launch

◦

Usage-based (typical spend ~$15/mo for intermittent load, 1 GB)/month

  • · Lakebase Postgres
  • · Managed Better Auth
  • · Object Storage (beta, no charges during beta)
  • · Functions (beta, no charges during beta)
  • · AI Gateway (beta, no charges during beta)
  • · Autoscaling

Scale

◦

Usage-based (typical spend ~$701/mo for high load, 100 GB)/month

  • · Lakebase Postgres
  • · Managed Better Auth
  • · Object Storage (beta, no charges during beta)
  • · Functions (beta, no charges during beta)
  • · AI Gateway (beta, no charges during beta)
  • · Autoscaling

Agent Plan

◦
  • · Designed for app generation platforms

Integraciones (12)

Next.jsPrismaDrizzleDatadogOpenAI SDKCursorClaude CodeGitHub CopilotpgvectorPostGISTimescaleDBVercel

Documentación

Badge para tu web

Muestra tu AI Readiness Score en tu web. Pega este código en tu sitio o footer; el badge enlaza de vuelta a este perfil.

AI Ready on Rootbase
HTML
<a href="https://rootbase.space/company/neon"><img src="https://rootbase.space/company/neon/badge.svg" alt="AI Ready on Rootbase" height="28"></a>
Markdown
[![AI Ready on Rootbase](https://rootbase.space/company/neon/badge.svg)](https://rootbase.space/company/neon)

Consultar por API / MCP

Este profile es consultable por agentes. Datos verificados el 2026-08-15 · versión 1.

GET  /v1/company/neon
GET  /v1/company/neon/pricing
POST /v1/query        { "company": "neon", "query": "..." }
MCP  /api/mcp         get_company · query_company · discover_companies

Fuentes