Payload CMS 4.0 is not just a release note. It is the first major public milestone since Payload joined Figma, and the moment the team finally shows the visible layer of a broader strategy. Here is a sourced early look: what is coming in 4.0, and what is being prepared "next", still largely unannounced.
At Beease, we already ship production sites on Payload + Next.js. We are very happy with the CMS as it stands: we are not waiting for 4.0 to "rescue" a stack. We are watching because the ecosystem is accelerating, and because a more polished admin will mostly help content teams.
Payload CMS 4.0 is not out (yet)
First thing to keep in mind: 4.0 is not stable. The official post frames it as an early look. Several pieces are already visible on the main branch of the official GitHub, but the team is clear: this is work in progress. Production still runs on 3.x.
Payload's own numbers since September: about 1,300 merged PRs, ~600 issues closed, ~181 features, and weekly downloads up from roughly 100k to 400k. The momentum is real, even if this is not a go-live yet.
The stated target: a beta in the next quarter, with a shorter beta-to-stable path than in 3.0 (where the April beta dragged into November). Paul Popus, on the team, puts it bluntly on Discord: first a public beta, then many releases between beta.0 and stable, with possible breaking changes based on feedback.
What is coming in 4.0
A redesigned admin (inspired by Figma's design system)
The most visible piece is a full admin redesign: lists and editing, tables, bulk actions, drawers, modals (crop / focal point), focus states, contrast mode, and cleaner Lexical controls. Sass gives way to semantic tokens, a real design system, and better coexistence with Tailwind when you customize the admin.
James Mikrut owns it in the community call: a lot of people complained about the UI for years, and it is partly on him ("maybe a little bit too minimalist").
Tylen Davis, Payload designer, sets the direction: a consistent experience between Payload and Figma, with a design system based on Figma's (
UI3). Clean Payload components, and a Figma library promised later.
Native hierarchies
Folders and tags become a core primitive, not just the nested docs plugin. The sidebar gets tabs (folders / tags), plus React slots for plugins (queues, favorites, multi-tenant…). Over time, nested docs will be replaced, even if the migration path is not locked yet.
For marketing sites with a real page tree, this is the kind of feature that changes day-to-day life for editors.
Agents, skills, MCP
Payload is shipping installable skills for agents (Payload-specific guides), LLM failure collection, and an evals suite. On the MCP side, the 4.0 goal is deliberately simple: add the plugin, configure MCP.json, and go. Tools are opt-out by default, and complex schemas become more reliable.
If you already follow the agents + CMS topic, also read our article on AI agents and Payload CMS.
TanStack Start (early)
Next.js remains first-class. But 4.0 introduces a framework adapters pattern (admin, RSC, loaders, API mounting). The first proof point: TanStack Start. The demo is early, CSS bugs are acknowledged, and the signal is clear: Payload no longer wants to be "Next-only forever", even if Paul had already cooled the rumor of a full decoupling as soon as 4.0.
Figma, Sites, and what is being prepared "next"
In June 2025, Payload joins Figma. James states the problem plainly:
"The gap between design and code still exists. Designers create in Figma, then devs recreate in code, then content teams struggle to maintain it all… And historically, the CMS tends to make it worse."
And the promise: integrate design systems "in ways no other CMS can".
Kris Rasmussen, Figma CTO, is even more direct on the Figma blog: when Sites was announced at Config, a CMS was already promised. Payload brings the backend developers love and an admin for editors. In other words, Sites CMS + Payload is not community speculation: it is the official framing.
Paul Popus (Payload team, now in the Figma ecosystem) says it straight on Discord: the ROI will be in what Payload does as a CMS inside the Figma platform, while open source stays a positive feedback loop. On the "rugpull" risk: technically impossible to pull everything back, the existing code is MIT.
Ry, Figma NYC designer, published the RFC Admin UX #13979 on September 30, 2025: he had already been working with James and the team for months. So the redesign is not a post-acquisition idea pulled out of a hat: it was already underway, quietly.
James, in the community call on v4:
"As we build these integrations back to Figma, we have a very wide array of ideas that we're actively building behind the scenes with Figma."
The admin redesign exists so the Figma ↔ Payload round trip feels coherent. It is not the final product: it is the visible prerequisite.
Discord anecdote (June 17, 2025): a user suggests "mockup Figma → convert → app CMS Payload end-to-end". James just replies: 😏. Not a roadmap, but probably the best unofficial teaser in the batch.
What we can see vs what stays in the dark
Short read, no science fiction. Short term: a Figma-aligned UI. Medium term: Payload as a serious content brain in, or next to, the Sites ecosystem. Long term (James's utopia): design systems and Payload models kept in sync, with less design / code / content handoff. Make, Dev Mode MCP, and agents can speed all of that up, but that is not the press release yet.
For broader CMS positioning, we already documented why Payload over WordPress. 4.0 does not change that foundation: it makes it more credible for editors, and more plugged into the Figma platform.
What we take away at Beease
- We are watching. No production on
main. 3.x remains the shippable line. - The CMS already works for us. The 4.0 UI is a plus for your marketing / content teams, not a technical rescue.
- The real "wow" features for our client projects: DAM, hierarchies, a clearer admin…
- Agents / MCP: clear direction, worth following closely. The simpler 4.0 setup could become a real differentiator.
- Figma: strategic context to know. When the product bridge goes public, we will talk about it with facts, not "😏".
Paul, again, cooling OSS fears: Figma has no interest in breaking open source. The stake is interoperability and the platform's CMS needs. OSS was "never" the paymaster: enterprise customers were. Blunt, useful, and consistent with what we have seen for a year: the open-source pace did not slow. It accelerated.
Sources and next steps
Primary sources to keep handy:
- Early look Payload 4.0 (Payload blog)
- Payload is joining Figma (James Mikrut)
- Payload joins Figma (Kris Rasmussen / Figma)
- RFC Admin UX #13979 (Ry, Figma)
- Community call on YouTube
- Press read: CMSWire on the Figma × Payload deal
- Payload Discord #general (June 17–18, 2025 announcement + May–June 2026 redesign teasing)
Payload CMS 4.0 is the public front. "And what's next" is closing the loop from design → structured code → content, with Payload as the source of truth in the Figma ecosystem. We do not have the product announcement yet, but we already have enough signals not to look away.
If you are building (or considering) a site on this stack, we can talk concretely: architecture, editor admin, and what 4.0 will change (or not) for you. Explore our Payload CMS agency offering, or get in touch.
Martin Lotz
CEO
Sales Engineer
Martin Lotz is CEO and Sales Engineer at Beease Digital. As a key member of the team, he bridges clients’ business needs with the agency’s technical solutions.
Schedule a demo with Martin
