The open data platform that wraps any SQL database with an instant API and a no-code app - no backend required.
In 2004, Ben Haynes was running a digital agency and doing the same thing over and over. Every new client meant a new project, and every new project meant building - yet again - an admin panel so the client could edit their own content. Slightly different flavors of the same backend, rebuilt from scratch, forever. It was the kind of repetitive work that quietly convinces an engineer there has to be a better way.
His answer was a reusable tool that could sit on top of any database and hand non-technical people a clean interface to manage it. The concept looked a lot like what the industry would later call a "headless CMS" - except the term didn't exist yet. This was years before the first iPhone. Haynes formalized the effort inside his studio, kept it open source, and let it grow.
The commercial company came much later. After a run of mergers and acquisitions landed Haynes in a leadership role at Verizon/AOL around 2015, he eventually left to focus on the project full-time. In 2020 he and Rijk van Zanten - who had modernized the framework as part of his master's work - incorporated Directus as a company. Haynes titled his own account of the journey honestly: "An overnight success, two decades in the making."
Most content and data tools ask you to migrate into their proprietary store. Directus takes the opposite approach. You point it at a SQL database you already have - PostgreSQL, MySQL, SQLite, OracleDB, CockroachDB, MariaDB or MS-SQL - and it introspects the schema and wraps it in two things at once: a real-time REST and GraphQL API for developers, and a no-code Vue.js admin app that non-technical teammates can use without training.
Nothing is hidden and nothing is locked away. Your data stays in your database, in tables you can still query directly. That single design decision is why Directus gets filed under three categories at once - headless CMS, backend-as-a-service, and no-code app builder - because in practice it does the job of all three.
Relative scale · approximate figures
Teams waste weeks writing CRUD APIs and admin dashboards for every project. Directus generates both instantly from an existing schema.
Engineers want a real API; marketers and ops want a friendly interface. Directus serves both from the same source of truth.
Proprietary platforms trap your content in their store. Directus keeps data in your own SQL database, exportable and queryable at any time.
Internal tools, dashboards and content often live apart. Directus unifies them over one database with roles, revisions and automation.
The open-source engine: instant REST + GraphQL API and a no-code Vue.js app over any SQL database.
Hosted projects with a free Community tier and paid plans from around $99/month.
Event-driven flows and operations for building data logic and automation without a separate service.
Install community and official extensions from inside your instance - no forking the core.
11 AI extensions for transcription, scraping, text intelligence and vision - with OpenAI and Anthropic support.
Run it yourself, free forever, with no artificial limits. Docker-ready and fully inspectable on GitHub.
The clearest way to understand Directus is by contrast. Where many competitors define their own data model and ask you to adopt it, Directus adapts to the database you bring. It is open source rather than closed, SQL-native rather than proprietary, and it serves developers and business users from the same layer.
| Dimension | Typical SaaS CMS | Directus |
|---|---|---|
| Data store | Proprietary, hosted | Your own SQL database |
| Source | Closed source | Open source, on GitHub |
| API | Vendor API only | REST + GraphQL, real-time |
| Hosting | Vendor cloud | Self-host or Directus Cloud |
| Lock-in | High | Take your data and leave |
| Users | Editors or developers | Both, same platform |
Alternatives people weigh Directus against: Strapi, Contentful, Sanity, Hasura, Supabase, Appwrite, Payload CMS.
The model is open-core. Self-hosting is free and open source, so individual developers and small teams pay nothing. Organizations with more than $5M in annual revenue or funding that run Directus in production need a paid commercial license, and Directus Cloud sells managed hosting on subscription. The free tier builds trust and community; the enterprise tier pays the bills.
Developers, digital agencies and large enterprises across financial services, consumer tech, travel and government. Named customers include Adobe, Bose, Tripadvisor, AT&T and the U.S. Navy - alongside a community of more than 50,000 developers who build with it every day.
Among the enterprises using Directus to power digital experiences and internal data tools.
Cited as a customer scaling content and data operations on the platform.
Listed among organizations turning to Directus for data management at scale.
Referenced as a Directus adopter, reflecting reach into government use cases.
Ben Haynes builds an early reusable admin panel at his agency - before "headless CMS" was a term.
After a run of mergers and acquisitions, Haynes lands in leadership at Verizon/AOL before choosing Directus.
Haynes and Rijk van Zanten incorporate Directus as a commercial open-source company.
Backed by Handshake Ventures, the team launches managed cloud hosting.
True Ventures leads a $7M round to build the industry's first open data platform.
Directus ships its Marketplace beta, releases v11 with policy-based access control, and rolls out AI extensions.
A monthly release cadence continues atop 30M+ downloads and a 50,000-strong community.
Directus lives in the composable, API-first corner of the software market - the same space as headless CMS and backend-as-a-service platforms, but anchored to open source and SQL. Its expertise is in database introspection, API generation, granular access control and a no-code interface polished enough that non-technical users need no training.
The 2024 release of Directus 11 rebuilt access control around a policies system, and the Marketplace turned the platform into a genuinely composable base others can extend. AI arrived not as a bolt-on chatbot but as a set of practical extensions - transcription, web scraping, text intelligence, image focal-point detection - wired into the existing automation engine. It is a company that absorbs trends into its data layer rather than chasing them.