We have been using Payload CMS on client projects since version 1.x, and the 3.0 release — which completed the move to Next.js App Router and added the new Lexical editor — was the update the project needed to become a serious alternative for certain use cases. Here is what we have learned after six months.
Developer Experience
Payload's developer experience is genuinely excellent. Config-as-code means your content schema lives in your repository, gets code-reviewed, and deploys with your application. You define your collections and access control in TypeScript, and the admin UI and REST/GraphQL APIs are generated automatically.
The new Lexical editor is a massive improvement over Slate. Rich text with embedded blocks — custom components rendered inline in the editor — works reliably and feels native. Our content teams adapted to it faster than we expected.
Next.js Integration
Payload 3.0 runs as a Next.js application, which means your CMS and front-end share a codebase, a deployment, and a runtime. Server Components can query Payload's local API directly — no HTTP round-trips, no API keys to manage, no CORS issues.
The Harder Parts
Payload is not a replacement for non-technical clients. The admin UI is clean but lacks the decade of UX polish that WordPress has accumulated. Expect to spend time customizing the admin experience for editorial teams. The plugin ecosystem is also thin — you will build more from scratch. For the right team, that is a feature. For teams that need to ship fast with less custom work, it is a real constraint.
Our Verdict
Payload CMS 3.0 is our default choice for new headless projects where we control the full stack. For TypeScript-native teams building modern web applications, it is the best CMS we have used.
