When to migrate from Framer: five structural limits worth understanding

Framer ranks among the best-designed tools for building marketing sites quickly, with design fidelity that no-code tools rarely match. It is designed for a specific use case, and when a company’s content volume, editorial team size, compliance requirements, or integration needs grow past that use case, the limits are structural. No plan upgrade removes them.

Framer 3.0 launched on June 16, 2026, adding AI Agents to the canvas, Git-style branching, and MCP support for external tools like Claude Code and Cursor. These are meaningful additions for design speed and team safety. They do not change the architectural limits that determine whether Framer is the right platform for a growing organisation.

What is the Framer ceiling?

Framer’s architecture makes a deliberate trade: visual design speed and design-to-publish continuity in exchange for a constrained CMS, per-seat editor pricing, limited approval workflows, and no content delivery API for external surfaces. That trade works for a large number of use cases and reaches its limits at a predictable set of thresholds.

The five signs below are the standard progression for companies that built correctly on Framer and then grew.

Sign 1: you are running into CMS collection and item limits

Framer’s CMS limits are plan-dependent and not as generous as they first appear. The table below reflects post-May 2026 pricing.

Plan	CMS collections	CMS items	Annual cost
Basic	2	1,000	$120
Pro	10 (expandable to 40)	2,500 (expandable to 40,000)	$360
Enterprise	Custom	Custom	Custom (contact sales)

The collection limit is the first one most teams hit, and it arrives faster than expected. A company with a blog, a resource library, a case study archive, team profiles, and a job listings section already needs five collections. Add localised variants, a changelog, or a partner directory and the Pro plan’s ten collections fill up. Additional collections are available as a paid add-on (up to 40 total), and the cost compounds as a content site grows.

The item limit is tighter than it appears. Pro starts at 2,500 items, a number that sounds manageable but is reached faster than most teams expect. A company with an active blog, a resource library, and localised variants of each can hit that ceiling within a year. Additional items are available via add-on (up to 40,000 total), but the base limit means growing content operations will likely encounter this constraint before the collection limit.

The Enterprise path exists but is custom-quoted and not published publicly. For a company that started on Framer because it was a lean, fast option, moving to a negotiated Enterprise contract changes the decision calculus significantly.

Sign 2: your editorial team is hitting the per-seat cost wall

Framer charges per editor seat on top of the site plan. Following Framer’s May 2026 pricing update, editor seats now cost $20 per month, down from $40 per month on Pro. A new Content Editor role at $10 per month covers CMS editing, localisation, and on-page editing for team members who do not need canvas access.

The reduction is meaningful. A three-editor agency on Pro now pays $70 per month rather than $110. For larger teams, though, the pattern still compounds. A marketing team of eight to ten people where most members need edit access adds $140 to $180 per month in seat costs above the plan cost on Pro. If those members only need content editing, the Content Editor role reduces that to $70 to $90, but the seats still multiply with headcount in a way that was not visible when the site was built.

The problem sharpens at scale. Payload carries no per-seat pricing on its open-source core. Contentful Enterprise negotiates seat allocations as part of the contract. Framer’s per-editor billing grows with every addition, regardless of role.

The per-seat model also creates a familiar incentive problem: organisations limit Framer access to manage costs, which means fewer people can publish content, which means the CMS that was supposed to give the marketing team editorial independence starts requiring a gatekeeper. The editorial velocity problem Framer was meant to solve reappears, driven by the pricing model.

Sign 3: you need approval workflows, audit trails, or compliance-grade access control

Framer has draft states for pages and CMS items, a staging environment, and a review step in on-page editing where editors without publish rights submit changes for a publisher to approve. Framer 3.0 adds Git-style branching, which lets teams make sweeping changes on a separate branch and merge to the live site only after review. For a small marketing team, this covers the basic need to review before publishing.

What the platform does not offer is a configurable multi-stage approval chain. Branching ensures design changes do not reach production before review. It does not provide named approvers per content type, role-based sign-off sequences, or a workflow where legal, compliance, and editorial must all approve before publication. For financial services, healthcare, legal, and other regulated industries, the distinction between “a publisher can review before pushing live” and “a defined approval chain with documented sign-offs” is the difference between a tool that fits and one that does not.

Framer also provides no structured audit trail. Version history for pages exists, but there is no queryable log of content changes that records who changed what, when, and what was published to production in a format compliance teams can export or attach to a regulatory filing. For organisations where content publication is a regulated activity, this is a concrete gap.

SSO is Enterprise-only. Single sign-on integration with Okta, Azure AD, Google Workspace, or OneLogin is available only on the Enterprise plan, which is custom-quoted. Organisations that require SSO for identity management or security policy compliance have no option within the Basic or Pro plan structures. When an editor leaves the organisation, their account must be manually deactivated by an administrator. On sites with frequent staff changes and large editor counts, this manual step becomes an access control gap that IT and security teams flag.

Sign 4: your content needs to go beyond the website

Framer’s CMS is designed to serve one surface: the Framer-hosted site. In March 2026, Framer launched a Server API (currently in open beta) that allows programmatic access to CMS data from external servers, enabling collection sync, publishing changes, and integration with tools like Notion or Airtable. Framer 3.0’s MCP support extends this further, letting external agents like Claude Code or Cursor import bulk content into the CMS or pull data from external feeds.

These are useful additions for content synchronisation and import workflows. What they do not provide is a content delivery API. A REST or GraphQL endpoint that a React Native mobile app, a partner portal, or an internal platform can query to retrieve and display published content (equivalent to WordPress’s /wp-json/wp/v2/) does not exist. The Server API is a management and sync interface. Framer community threads from 2024 through 2026 show users consistently requesting content delivery capability, which signals the gap remains open.

Mobile applications. A React Native mobile app that needs to display the same product descriptions, help articles, or onboarding content as the marketing site cannot query Framer’s CMS as a content source. The content would need to be duplicated and maintained separately in a system the mobile app can access. Two sources of truth for the same content is a governance problem that compounds with every update.

Partner portals and authenticated web applications. Applications that require authentication or serve content to specific user segments cannot use Framer’s hosted pages. Serving personalised or gated content to different audiences requires a CMS that exposes a content delivery API.

Internal platforms. Intranets, operations dashboards, and internal knowledge bases that need to surface the same structured content as the public site cannot connect to Framer’s CMS as a data source. Internal content must be managed separately, creating the same duplication and governance problem as the mobile case.

A headless CMS delivers content through an API that any surface can consume. Payload, Contentful, and Sanity all work this way. The web application, the mobile app, the internal portal, and the partner extranet all query the same content API and render the content for their surface. Framer’s CMS architecture does not support this model.

Sign 5: you cannot take your design system with you

Content is exportable via the official CMS Export plugin (CSV) or a JSON Import and Export plugin. Framer’s documentation explicitly supports content portability, partly in response to EU Data Act requirements effective September 2025.

The CSV export covers text fields and metadata. Rich text fields export as HTML that requires transformation. Image references export as URLs that need to be re-ingested into the destination platform, reference fields between collections require custom handling. The export exists, but migrating to a headless CMS from it still requires ETL scripting rather than a direct import.

The design system has a different problem: there is no export path. Framer’s component system and interactions are proprietary. Animations built in Framer’s interaction editor do not export to React components or CSS that another system can consume. Moving to a Next.js or Astro front end means rebuilding animations using Framer Motion (the React library) or other animation tooling. The visual output can be reproduced; the production mechanism cannot be ported.

This is an architectural property of how the platform is built, and it is worth understanding before it becomes a migration constraint. For a company making a multi-year platform investment, non-portable design tooling creates a rebuild cost that compounds with the size of the design system.

The alternative is a composable stack: a headless CMS such as Payload or Contentful for content management, and Next.js or Astro for rendering. Content lives in a database the organisation owns, accessible via a public API. The codebase is standard TypeScript. Every piece is portable. The migration path from this stack to any other destination is an engineering project with a defined scope.

The Framer decision framework

Hitting one of these signs is not itself a reason to migrate. The signs are indicators of when migration becomes worth evaluating.

Are you hitting CMS collection or item limits on Pro?
├── No  → Stay on Framer; revisit in 12 months
└── Yes → Is your editorial team larger than 4–5 people?
           ├── No  → Evaluate based on content model needs (Sign 4)
           └── Yes → Do you have approval workflow, SSO,
                      or compliance requirements?
                      ├── Yes → Migration case is strong;
                      │         evaluate Payload or Contentful
                      └── No  → Does your content need to serve
                                  surfaces beyond the Framer site?
                                  ├── No  → Pro add-ons or Enterprise
                                  │         may still be justified
                                  └── Yes → Migration case is strong;
                                            evaluate Payload

For organisations where three or more of the five signs are present simultaneously, the migration case is typically clear. For organisations where one or two signs are present, the decision depends on how quickly the remaining thresholds are approaching.

WAYF builds Framer sites and migrates organisations off Framer when they have outgrown it. The hybrid path, keeping Framer for design-intensive brand pages while routing structured content through Payload, is also an option when the rendering layer is not the constraint.

---

FAQ

1. What are Framer's CMS limits?

   As of June 2026, Framer offers three paid plans. Basic includes 2 CMS collections and 1,000 items (updated from 1 collection in May 2026). Pro includes 10 collections and 2,500 items, expandable with add-ons to 40 collections and 40,000 items. Enterprise plans negotiate custom limits. There is no Scale plan. The collection cap is typically the first limit growing companies hit, as ten collections fill quickly when a site includes a blog, resource library, case studies, team profiles, job listings, and localised variants.

2. Does Framer have a public content API?

   Framer launched a Server API in March 2026 (open beta) that allows programmatic access to CMS data from external servers for sync and management purposes. Framer 3.0 added MCP support, enabling external agents like Claude Code or Cursor to import and organise CMS content. What Framer does not have is a content delivery API: a REST or GraphQL endpoint that a mobile app, partner portal, or internal platform can query to retrieve and display published content. Organisations that need to serve Framer CMS content to external surfaces do not currently have a supported path to do so.

3. Does Framer support SSO?

   SSO with providers including Google Workspace, Azure AD, Okta, and OneLogin is available only on Framer's Enterprise plan, which is custom-quoted and not published publicly. There is no SSO option on the Basic or Pro plans. Organisations that require SSO for identity management or security compliance must look to other platforms.

4. Can you export content from Framer's CMS?

   Yes. Framer provides a CMS Export plugin that exports any collection as a CSV file, and a JSON Import and Export plugin for JSON format. Framer's documentation explicitly supports content portability, partly in response to EU Data Act requirements. The export covers text fields and metadata, but rich text exports as HTML requiring transformation, and image references need to be re-ingested into the destination platform. A clean migration to another CMS still requires ETL scripting rather than a direct import.

5. Does Framer have approval workflows?

   Framer has draft states, a staging environment, and a review step in on-page editing where editors without publish rights submit changes for approval by a publisher. Framer 3.0 added Git-style branching, which lets teams make changes on a separate branch and merge to the live site only after review. For small teams, this covers the basic need to review before publishing. Framer does not have a configurable multi-stage approval chain: named approvers per content type, role-based sign-off sequences, or compliance-grade workflow with documented, exportable audit trails. For regulated industries where content approval is a documented, auditable process, this gap is material.

6. What changed with Framer 3.0 and the May 2026 pricing update?

   Framer 3.0 (June 2026) added AI Agents to the design canvas, Git-style branching for safer team collaboration, and MCP support for external AI tools. The May 2026 pricing update reduced editor seat costs from $40/mo to $20/mo on Pro, introduced a new Content Editor role at $10/mo (CMS, localisation, and on-page editing access, no canvas), and upgraded the Basic plan's bandwidth from 10GB to 50GB and collections from 1 to 2. These updates improve Framer for the use cases it was already suited to. The structural limits covered in this article, CMS collection caps, no content delivery API, no compliance-grade approval workflows, and non-exportable design tooling, are unchanged.

7. When should you migrate from Framer to Payload?

   Migration from Framer to Payload is justified when three or more of the following are true: hitting CMS collection or item limits; editorial team larger than four or five people with per-seat costs compounding; compliance-grade approval workflow, SSO, or audit trail requirements; need to serve content to surfaces beyond the Framer site such as a mobile app, portal, or internal tools; or design system complexity that requires a portable, code-based front end. If only one condition applies, Pro add-ons or Enterprise may extend the runway. If three or more apply, the migration case is usually clear.

8. What is the hybrid Framer and Payload approach?

   The hybrid pattern keeps Framer as the rendering layer for design-intensive brand and marketing pages, while routing structured content through Payload as a headless CMS. Framer custom code components fetch content from Payload's API and render it within the Framer site. This preserves Framer's visual design capabilities for pages where design fidelity matters most, while providing a content API, approval workflows, and CMS export capability for content that has outgrown Framer's native CMS.

9. How much does Framer Enterprise cost?

   Framer Enterprise pricing is custom-quoted and not published publicly. It is negotiated per organisation based on limits, SSO, seat counts, and support requirements, so expect a tailored annual contract rather than a list price.

Book a call

---

Sources

• Framer: Pricing Plans
• Framer Blog: AI credits, simpler plans, and lower prices
• Framer Blog: Introducing Framer Agents, Branching, and the new Community
• Oma-kase: Framer Pricing Updates May 2026* All About Framer: Framer’s New Pricing: What It Really Means in 2026