☁️Build Your Own Cloud

Vercel hosts frontend applications with Git-based deploys, preview URLs, edge functions and first-class Next.js support.

// TAKE · Coolify is the most complete self-hosted answer to Vercel, giving you Git-triggered builds, preview deployments, managed databases and certificates from a single UI. Dokploy is a lighter, faster-moving take on the same idea and Dokku is the dependable minimal choice for one VPS, but none of them reproduce Vercel's edge network or its Next.js-specific build optimisations.

Netlify hosts static sites and serverless functions with Git-based deploys, preview builds, an edge CDN and form handling.

// TAKE · Coolify is the closest self-hosted Netlify: Git-push deploys, preview environments, automatic TLS and databases behind one dashboard. Dokploy covers similar ground with a lighter Docker and Compose focus, Dokku is the minimal, battle-tested option if you are happy in a terminal, and Stormkit mimics the frontend workflow most literally but has a much smaller community.

Heroku runs a managed platform that builds and hosts apps from a git push, with add-on databases, dynos and easy scaling.

// TAKE · Dokku is the most faithful to Heroku itself, using the same buildpacks and git-push deploys, though it remains a single-server tool. Coolify and Dokploy give you a modern web UI, managed databases and multi-server deploys on top of Docker, in exchange for a heavier stack to keep running.

Sevalla is a managed platform for deploying applications, databases, static sites and object storage without running your own servers.

// TAKE · Coolify is the closest like-for-like, covering git-driven app deploys, managed databases and one-click services on hardware you own. Dokploy handles the same ground with a cleaner Docker Compose story, while Dokku is the minimal choice if you want one server and are happy driving it from the CLI.

Supabase provides a hosted Postgres backend with auth, instant REST and realtime APIs, file storage and edge functions.

// TAKE · NHost is the nearest match on architecture, pairing Postgres with auth, storage and generated APIs, while Appwrite covers the same surface with its own datastore and a more polished console. PocketBase trades Postgres for a single SQLite binary you can run on a small VPS, which is a genuine win for side projects and a genuine ceiling for write-heavy ones.

Firebase provides a hosted app backend with realtime and document databases, authentication, file storage, push and serverless functions.

// TAKE · Appwrite is the closest like-for-like, matching Firebase's auth, storage, functions and realtime APIs with client SDKs for most platforms. PocketBase does the same job in a single Go binary when your scale fits on one machine, and Convex or Instant are the better picks when live-syncing state to clients is the whole point of the app.

AppWrite is a backend-as-a-service platform bundling auth, databases, file storage, serverless functions and messaging behind SDKs.

// TAKE · PocketBase is the closest drop-in: one binary with auth, a database, file storage and realtime subscriptions, though it is SQLite-backed and scales vertically only. Bknd covers similar ground with a framework-agnostic embed story, while Encore and Nitric solve the opposite problem, generating cloud infrastructure from your backend code rather than handing you a ready-made one.

PlanetScale is a managed MySQL platform built on Vitess, offering horizontal sharding and branch-and-merge workflows for schema changes.

// TAKE · TiDB is the closest technically, a MySQL-compatible database that shards horizontally without application changes, though nothing in it resembles PlanetScale's branching schema workflow. Turso fits a different shape of problem well, thousands of small per-tenant SQLite databases rather than one large cluster, and CrateDB leans analytical rather than serving as an OLTP replacement.

MongoDB is a document database storing JSON-like records with flexible schemas, sharding and the managed Atlas cloud service.

// TAKE · Nothing here is a drop-in for MongoDB's wire protocol, so choose by what you actually used it for. Convex and Instant come closest to the developer experience of schemaless documents with live sync to clients, TiDB is the move if you outgrew Mongo and want distributed SQL with strong consistency, and Memgraph only fits if your data was really a graph all along.

Amazon S3 is AWS object storage offering durable buckets, lifecycle tiering and the S3 API that most storage tooling now speaks.

// TAKE · MinIO is the closest drop-in, compatible enough that most SDK code needs only an endpoint change, though its licensing and feature gating have pushed some teams to look elsewhere. Garage is the better fit for small self-hosted clusters spread across a few locations, and Ceph only earns its complexity if you need block and file storage from the same cluster.

Azure Blob Storage is Microsoft's cloud object storage for unstructured data, with hot, cool and archive tiers and HTTP access to blobs.

// TAKE · MinIO is the practical swap, but note it speaks S3 rather than the Azure Blob API, so budget for a client rewrite before you compare anything else. Ceph is the choice when you need block and file storage next to objects and can staff the operational load, while Garage is the sensible option for a small multi-site cluster on modest hardware.

Google Cloud Storage is Google's object storage service for buckets of files, with storage classes, lifecycle rules and global access.

// TAKE · MinIO is the default answer since everything already speaks S3 and it ships as a single binary, though licensing and feature gating have pushed some teams to Garage, which runs happily on small, geographically spread nodes. Ceph is the serious option when you also need block and file storage from one cluster but demands real operational effort, and SlateDB is not a replacement at all, it is a storage engine that sits on top of object storage.

Terraform declares cloud infrastructure as code and provisions it across providers using state files and a plan/apply flow.

// TAKE · OpenTofu is the answer for almost everyone leaving Terraform over the BSL licence change: it is a fork, so your HCL, modules and state files carry over with minimal edits. Pulumi is the bigger change and worth it only if you genuinely want infrastructure in TypeScript, Go or Python; Digger and Terrateam do not replace Terraform at all, they replace the paid Cloud layer by running plan and apply in your own CI with PR approvals.

Red Hat Ansible automates server configuration, application deployment and IT orchestration using agentless, YAML-defined playbooks.

// TAKE · None of these is a direct configuration-management swap, but OpenTofu comes closest for declarative infrastructure you can version and re-apply, and it is the safest choice given the community behind it. Pulumi expresses the same thing in real programming languages if your team prefers code over HCL, while Harness covers the deployment pipeline rather than the machine state.