Contentful alternatives, sorted by what you are leaving for

When we read the first page of results for “Contentful alternatives” on 1 September 2026, every result was a vendor explaining why its own product was the answer. Results move and they vary by searcher, so check yours. The pattern is the thing worth noticing: a ranked list published by a company that sells one of the entries tends to open at number one with that entry.

This page asks the question we put to a team that tells us it is looking at leaving: what are you actually leaving for? The answer sends you down one of three branches, and the candidates that fit one branch are the wrong candidates for another.

Where we stand, before you read any of it

WAYF is a Payload partner agency and top contributor. Payload appears in one branch below. We have a commercial interest in you picking it, and you should read the Payload paragraphs knowing that.

So the page is built to be checked. Every pricing, hosting and licensing claim below links to the vendor’s own page and carries the date we read it. Where a page would not load for us, we said so instead of estimating; in the end every figure below came from a vendor page. And the branch where Payload is the wrong answer is written as plainly as the branches where it fits, because a recommendation that never says no is a brochure.

Last verified: 8 September 2026. Vendor pricing pages change without notice. Check the linked source before you build a business case on any figure here.

The question that sorts everything

Teams leave Contentful for one of three reasons, and they rarely say which one out loud.

1. The bill climbs with usage. Users, roles, locales, API calls and CDN bandwidth step with the plan. Content types, environments and records are quotas attached to each Space, and Spaces are bought separately.
2. The platform is a poor fit and you still want somebody else to run it. The bill is survivable. The modelling, the editor, or the developer experience is what has stopped working.
3. The Salesforce acquisition changed your risk assessment. Ownership moved on 1 September 2026 and you are deciding what that means for a renewal.

Work out which one you are before you shortlist anything. The three lists barely overlap.

What you are comparing against

These figures come from Contentful’s pricing page, read on 8 September 2026. Free is $0 and carries 10 users, 2 roles, 2 locales, 100K API calls and 50 GB of CDN bandwidth a month, on hard caps with no overages. Lite is $300 a month for 20 users, 3 roles, 3 locales, 1M API calls and 100 GB, and it does allow paid overages, at $5 per additional million API calls and $0.15 per additional GB of bandwidth. Enterprise is quoted, with API calls listed as waived.

Space quotas sit underneath the plan and are where the step gets steep. The Starter Space included on every plan carries 25 content types, 2 environments and 10,000 records, with workflows unavailable. The Lite Space that raises those to 50 content types, 4 environments, 50,000 records and up to 5 workflows costs $850 a month on top of the $300. So a single-project team that outgrows 10,000 records is looking at $1,150 a month before it has added a second project.

One caveat on reading the numbers. Contentful’s technical limits page publishes far higher ceilings — 5,000,000 records per Space, 500 locales per environment, 5,000 users per organisation — and says on the same page that where your plan or Space entitlement is lower, the entitlement is what binds.

Branch one: the meter is the problem

If usage-based pricing is what put this on your roadmap, the move that changes the shape of the bill is the move to software you run yourself. The licence line goes to zero and the spend becomes infrastructure and engineering time. Neither is fixed. Database, object storage, egress and compute all move with traffic and content, and maintenance moves with how much you have built. What changes is the shape: you stop paying per record, seat and API call, and start paying for what your application consumes and what your team spends keeping it healthy. Whether that is cheaper at your volume is an arithmetic question about your own numbers.

Payload. The licence file is the plain MIT text, and GitHub reports the repository’s licence as MIT. It deploys anywhere Next.js runs — the documentation names Vercel, Netlify, SST, DigitalOcean and AWS — and the same page is candid that most projects also need a database, file storage, an email provider and a CDN. The content model is TypeScript in your repository, so a schema change is a pull request. That deployment page describes self-hosting throughout and offers no managed alternative on it, so plan on the operational work landing with your team.

Strapi. Also self-hostable, with a licence that needs reading carefully. The repository licence grants the Community Edition under MIT Expat, and puts anything under an ee/ directory under a separate Enterprise licence. GitHub reports the project as NOASSERTION for that reason. If you are self-hosting to escape a vendor relationship, check which features you need sit on which side of that line.

Strapi Cloud runs the same software hosted. As of 8 September 2026, Starter is $35 per month for 100k API requests and 50 GB of asset storage, Pro is $90 for 1M requests, and Business is $450 for 10M requests with a 99.9% uptime SLA. Overages are billed at $1.50 per 25k requests and $30 per 100 GB of bandwidth, and prices exclude local taxes. Read the Starter row before you price a production site on it: the page lists it as sleeping when idle, against “always on” for Pro and Business, and shows no backup frequency where Pro carries weekly backups and Business daily, both retained a month. Give Strapi the point here. Every Cloud tier lists database entries as unlimited, including the $35 Starter, where the hosted platforms in the next branch meter content by story, document or entry and raise the allowance as you move up the price list. Strapi answers “this category of software, hosted, at a published price” in a way we could check, and Payload does not: its pricing page returned 404 on 8 September 2026 and its cloud page redirects to a login, so we have no comparable published tier to set beside it.

Where this branch goes wrong is capacity. Self-hosting moves a recurring invoice onto a team that has to keep a database, a CDN and a deployment pipeline healthy. If that team does not exist, or exists and is already full, you have swapped a predictable cost for an unpredictable one. Say so at the shortlist stage rather than three months into a build.

Branch two: stay hosted, change vendor

If you want a vendor on the hook for uptime and upgrades, the question narrows to which hosted platform meters what you consume. All three below publish self-serve prices and quote their enterprise tier separately.

Sanity. The free tier is the unusual one in this group. Sanity’s pricing puts it at up to 20 user seats with 2 permission roles, 2 public datasets, and unlimited content types and locales. Unlimited there has a ceiling underneath it. The page caps unique attributes per dataset at 2,000 on Free and 10,000 on Growth, and its FAQ says you can create as many content types as you like while you stay inside that allowance; locales draw on the same pool. The usage quotas bind separately: 10,000 documents, 250k uncached API requests, 1M CDN requests, 100 GB of assets and 100 GB of bandwidth a month. Growth costs $15 per seat per month, allows up to 50 seats and 5 roles, and lifts the document count to 25,000; the page states that viewers are free and do not count toward seat usage. Read the roles before the seat count, though. The two on the free tier are Administrator and Viewer, so anyone who edits is an administrator; Editor, Developer and Contributor arrive with Growth. Twenty seats at no cost is a real allowance, and it comes without a way to let someone write without also handing them the keys.

Storyblok. Its pricing starts free at one space, one seat, 2 locales, 100k API requests and a 20,000-story cap, with a second seat at $15 a month. Growth is $99 monthly, or $90.75 a month on annual billing, includes 25,000 stories, and it still includes 2 locales; additional ones are $20 a month each. Growth Plus at $349 includes 100,000 stories against an unlimited maximum, so read that $349 as the allowance rather than the ceiling. If multilingual is why you are leaving Contentful, price the locales before you price the plan.

Hygraph. Tightest free tier of the three on content volume: its pricing lists 1,000 entries against 3 seats, 2 locales and 500,000 API calls. Growth, priced from $199 a month, takes that to 10,000 entries, 10 seats and 3 locales, with API overages at $0.20 per million operations and asset traffic at $0.20 per GB. The locale and model ceilings lift only at the enterprise tier, which the page puts at 80 locales and 500 models.

Read those caps against your own numbers before the feature comparison. A platform that fits your workflow and caps out on entries in year two is a migration you will run twice. And check them against Contentful’s rather than assuming a move is an upgrade: Contentful’s free plan carries 10 users, so on seats alone Sanity is the one of these three that clearly beats it, and Storyblok starts at one.

Branch three: staying is a real answer

If the Salesforce acquisition is what moved this up your list, start from the read we published at the close: most Contentful owners will stay. Staying is a defensible outcome on its own terms.

The close transferred ownership of the company. It did not transfer your terms: entitlements, prices and support commitments run as written until your term ends, and Contentful’s published assignment language already permitted transfer to a successor. At close there was no Contentful entry on Salesforce’s active product retirement register. We wrote up what the close changed and what it did not in what changes today, and the checks worth running are in what to do while the deal is open.

Your calendar decides your position. A renewal or notice-to-non-renew deadline inside the next two quarters is the condition that moves you from watching to assessing. Further out than that, you have time to run an assessment properly, and running one badly under time pressure is how teams end up on a platform that fits worse than the one they left. One deadline is independent of the acquisition: Compose’s end-of-2026 date was set before the close, so Compose-dependent teams are on that clock either way.

What the move costs

Every branch that ends in a migration carries the same kind of bill, and the licence saving is the easy half to calculate.

Scope the question to the tool you will actually use. These are limitations of the standard CLI export and import path. Data that path leaves behind may still be reachable through other APIs, at the cost of writing the code to fetch it. Contentful’s CLI import and export documentation sets out what that path leaves behind: versions, releases, tasks and workflows, space memberships, changes to author history, custom apps, and webhook secrets. Roles are the exception worth knowing about. The export library behind that path does carry them: its output includes a roles array and its skipRoles option defaults to false, so roles come across unless you turn them off. Its introduction still says roles and permissions will arrive “in a future version”, which the options and the output shape contradict. Space memberships and webhook secrets are separate items, and both sit in the tool’s own limitations list.

Two further items come from our own account of the close rather than from Contentful’s CLI documentation: SSO configuration and the GraphQL schema registry. Treat those as separately sourced and check them yourself. Each item on both lists is a decision: rebuild it, replace it with a process, or accept losing it. Run an export dry run and diff what comes back against what your team depends on, because the lists move as the tooling changes.

The transform has three predictable cost centres. Assets move before entries, because upload and relationship fields need media IDs to point at. References need a two-pass import: create every document with relationship fields empty while recording an entry-ID-to-document-ID map, then resolve the references once the targets exist. Rich Text is a separate JSON tree rather than markup, so it is transformed into whatever shape the destination editor stores, with embedded entries resolved after IDs exist. Locales are remapped explicitly. The full technical path is in migrating from Contentful to Payload.

Operating effort is the part that outlives the project, and neither model reduces it to nothing. Hosted moves the CMS’s uptime, patching and infrastructure to the vendor. Your team still owns the front end, the schema, the integrations, the delivery pipeline and the content operations, and it still owns watching the plan as usage climbs. Self-hosted keeps all of that and adds the database, object storage, backups, deployment and an upgrade cadence for the CMS itself. Payload’s deployment documentation is a list of the services the application needs, which is a scope of work rather than a promise about cost. Price both against the team you actually have, over three years.

Where we would tell you not to pick Payload

A large editorial group with no engineering capacity should stay on hosted software. Payload’s own deployment documentation puts the database, file storage, email provider and CDN on your team, and describes no managed alternative on that page. If your organisation cannot name the person who will apply the next security update, the self-hosted branch is the wrong branch, and the fact that we build on Payload daily does not change that.

Mature roles and editorial workflow are the other case. Where a large team already depends on a hosted admin with granular permissions and approval steps, the honest comparison runs against another hosted platform. We set that out feature by feature in Contentful vs Payload.

Sources

Every figure above was read at the vendor’s own page on 8 September 2026:

• Payload licence and deployment documentation
• Strapi repository licence and Strapi Cloud pricing
• Sanity pricing
• Storyblok pricing
• Hygraph pricing
• Contentful pricing and technical limits

Every price above is quoted in US dollars as the vendor pages present them. Pricing pages vary by region, so the figure you see may differ.

One authority outranks all of this for your own account: your contract. Entitlements, prices and support commitments run as written until your term ends, whatever the public page says this week. Check the numbers here against your Service Order before you build a business case on them.

If you want this assessment run against your own usage, tell us what you are running. The first conversation is free, and a scoped fixed quote comes after it.

FAQ

1. What are the main Contentful alternatives?

   They divide by what you are leaving for. If the problem is cost that climbs with usage, the self-hosted candidates are Payload and Strapi, where the software is free and the spend becomes hosting and engineering. If the problem is Contentful specifically and you want to stay on hosted software, the candidates are Sanity, Storyblok and Hygraph. If the problem is the Salesforce acquisition, staying is a real option, because the close transferred ownership of the company rather than the terms you signed.

2. Is there an open-source Contentful alternative?

   Two, with different licences. Payload's licence file is the plain MIT text and GitHub reports the repository's licence as MIT. Strapi is dual licensed: its Community Edition is under the MIT Expat licence, and anything under an ee/ directory in the repository is under a separate Enterprise licence, which is why GitHub reports the project's licence as NOASSERTION rather than MIT. If your procurement process requires a single permissive licence over the whole codebase, that difference is the one to check first.

3. Is there a free Contentful alternative?

   Free-to-licence and free-to-run are different questions. Payload's software is MIT licensed at no cost, and Payload's own deployment documentation says most projects also need a database, file storage, an email provider and a CDN, so the bill moves to infrastructure and engineering. Among the hosted platforms, Sanity, Storyblok and Hygraph each publish a free tier with caps: as of 8 September 2026 Sanity's free tier covers up to 20 seats and 10,000 documents, though its only roles there are Administrator and Viewer, so everyone who edits is an administrator; Storyblok's covers one space, one seat and 20,000 stories; and Hygraph's covers three seats and 1,000 entries. Contentful itself has a free plan on the same page, at 10 users, 2 locales, 100K API calls a month and a Starter Space of 10,000 records, so compare against that rather than assuming any move off it saves money at the small end.

4. Should we leave Contentful because Salesforce acquired it?

   The acquisition on its own is not a reason to move. The close transferred ownership of the company, and entitlements, prices and support commitments run as written until your term ends. What decides your position is your renewal date and your notice-to-non-renew deadline. A renewal inside the next two quarters moves you from watching to assessing; anything further out leaves time to run the assessment properly.

5. What does it cost to move off Contentful?

   The licence saving is the easy half to calculate and the smaller half of the decision. The work is in the export, and the limits belong to the standard CLI export and import path rather than to every API. Contentful's CLI import and export documentation sets out what that path leaves behind: versions, releases, tasks and workflows, space memberships, changes to author history, custom apps, and webhook secrets. Two more items, SSO configuration and the GraphQL schema registry, come from our own account of the close rather than from that documentation, so check them separately. Rich Text is a JSON tree that has to be transformed into the target editor's format, and references need a two-pass import. Run an export dry run and diff the result against what your team depends on before anyone quotes you a number.

6. When is Payload the wrong choice?

   When nobody on your side is going to run it. Payload's deployment documentation puts the database, file storage, email provider and CDN on your team, and describes no managed alternative on that page. A team with a large editorial group and no engineering capacity is better served by hosted software, and that is true whether the hosted vendor is Contentful, Sanity, Storyblok or Hygraph. We are a Payload partner and this is still the answer we give.