Sanity in 2026: The Content Operating System
Most content management systems give every business the same editing screens and ask it to fit its content into them. That works until the content stops looking like pages: product details shared by a website, an app and a shop, a guide published in six languages, or copy that AI tools now need to read as well as people.
Sanity takes the opposite approach. It treats content as structured data and lets your developers build the editing environment around the way your team actually works. Sanity counts more than a million users and over 6,000 teams, and the customers it names include Figma, Shopify, Spotify, Nordstrom and PUMA. This guide explains what it is, how it handles content federation and how to tell whether it fits.
What Sanity is
Sanity describes itself as the Content Operating System for the AI era. In practice it has three parts. The Content Lake is a hosted database built for content, where every document is stored as structured data. Sanity Studio is the editing environment, open source under the MIT licence, which developers configure in code and can host wherever they like. And a set of APIs delivers the content to websites, apps and, increasingly, AI agents.
The company began in Oslo, Norway, was in production with customers from 2015 and launched publicly in 2017. It now works from San Francisco and Oslo, is led by co-founder and CEO Magnus Hillestad, and raised an $85 million Series C in May 2025, led by GP Bullhound.
GROQ in plain terms
GROQ is Sanity's own query language, open source and published as a full specification. A website or app uses it to ask for exactly the content it needs in one request. A product page can fetch the name, the images, the reviews and the related articles together, follow links from one document to another, and get the result back already shaped for the page. Content is served through Sanity's CDN, and the Live Content API, available on every plan, can update what visitors see as soon as a change is published.
Sanity also offers GraphQL, but as a generated, read-only layer: it can fetch content but not change it, and Sanity itself warns that a content model may need adjusting before a GraphQL API can be deployed. A team set on GraphQL should weigh that, and a team new to Sanity should allow time to learn GROQ.
Content federation
Content federation means making content that lives in several systems available through one API, without copying or migrating it. It is the idea behind Hygraph's Remote Sources, which fetch data from other platforms at the moment a page asks for it, as our Hygraph guide explains.
Sanity takes a deliberate position here. It aims to be one authoritative, well-modelled source of content rather than a gateway to other systems' APIs, and it offers three ways to work with data that lives elsewhere. A field can hold another system's ID, such as a product's, and your website combines it with live values from that system, like price and stock, when the page is built. Where a copy is really needed, connectors and Functions keep one in sync: Sanity Connect for Shopify, for example, adds your products to the Content Lake as documents editors can enrich. And on Enterprise plans, cross-dataset references link content between datasets within one Sanity project.
The practical difference is where the joining happens. Hygraph does it inside the CMS when a page asks for content; with Sanity it happens ahead of time through a sync, or in your own code. A synced copy keeps pages working when the source system is slow, but it is only as fresh as the last sync. A value looked up by its ID is always current, but your site then depends on that system being available. Neither is right in every case, which is why mapping your data sources comes before choosing a platform.
What editors and developers get
Editors work together in real time in Sanity Studio, and with the Presentation tool they can click on a live preview of the website and edit the content behind it. Comments, tasks and scheduled drafts come with the Growth plan, and certain Enterprise plans add Content Releases, which schedule changes across many documents at once. Content can be localised field by field or as whole documents, and AI features such as AI Assist and Canvas help with drafting and editing.
Developers define the content model in JavaScript or TypeScript, so it lives in version control alongside the website's code. They also get webhooks that fire when content changes, Functions that run custom logic in response, an App SDK for building custom editorial tools, and an MCP server that connects AI assistants such as Claude and Cursor to the content.
Where it asks more of you
The Content Lake is hosted only. Sanity Studio is open source and can run on your own hosting, but the content itself always lives on Sanity's platform, so plan how you would export it from the start. Sanity states that it runs on Google Cloud, is SOC 2 Type 2 certified and is GDPR and CCPA compliant, and it lists Belgium as its primary hosting region.
Sanity is a developer's tool first. Nothing is ready out of the box until someone has modelled the content and set up the Studio, which is the source of its flexibility and also of its setup cost. Each plan includes usage quotas, and usage beyond them is billed, so it pays to estimate your traffic early. Governance features such as single sign-on, custom roles and the full audit trail sit on the Enterprise plan, so check early which of them you will need.
When Sanity is the right choice
Sanity is a strong fit when one well-structured source has to serve many channels at once, such as a website, an app, a shop and AI tools, when the editing experience should follow your team's workflow rather than the other way round, and when you have developers comfortable with JavaScript who can own the setup.
It is a weaker fit when the main job is to combine live data from many systems at query time, which is exactly what Hygraph is built for, and for teams that need to host their content themselves. For those, an open-source option such as Payload is worth a look.
How we approach a Sanity build
We start with the content model, designed around reuse so the same content can serve every brand, market and channel, and we settle the localisation strategy before any content is entered. For each outside system we decide deliberately whether its data is referenced by ID or synced as a copy, and how fresh each piece needs to be. We shape Sanity Studio around how your team works, with roles, previews and Visual Editing connected to the live site, serve published content through the CDN, and set up webhooks or Functions so the site updates as soon as something is published.
Considering Sanity for a new website, an app or a multi-brand build? Talk to us. We will map your content and data sources and tell you plainly whether it is the right fit.
Sanity and the Sanity logo are trademarks of Sanity. This guide is independent and is not sponsored or endorsed by Sanity.
