Everything you need to run content. Nothing you have to rent.
A complete content backend that is code-first, type-safe, self-hosted, and yours. Here is the whole thing, feature by feature.
Ship changes without holding your breath.
Most CMSs charge extra for a second space or a staging environment. Tangen treats both as core. Model your content once, then run it across as many workspaces and environments as your team needs.
Many workspaces, one install
Run isolated content workspaces side by side for different brands, teams, or clients. Each has its own schema, data, members, and roles.
Real environments
Development, staging, and production are first-class and data-isolated. Test schema and content changes safely before they reach live.
Per-environment everything
Config, extensions, routes, and API keys are scoped per environment. What runs in staging can differ from production, on purpose.
One workspace, three environments
Data-isolated. Promote schema and content forward when it is ready.
# push schema to a specific environment $ tangen push --env staging $ tangen push --env production # every env has its own data, keys, and config
Your schema is real TypeScript.
Define object types, fields, relations, blocks, and taxonomies in code. Tangen generates the database, the API, and a typed SDK from it, so your model and your types never drift apart.
- Fields that fit real content. Text, rich text, relations, blocks, repeaters, media, and custom field types you author yourself.
- Generated types. Every type you define flows into the SDK and admin, so a wrong field is a compile error, not a runtime surprise.
- A visual builder too. Prefer clicking? The visual schema builder writes the same code, so design and engineering stay in sync.
Authentication, built in.
Members, sessions, email verification, and API keys ship with the platform. There is no separate auth service to wire up and no third-party redirect dance for your own users.
- Members and sessions. Invite teammates, verify emails, and manage sessions out of the box.
- API keys, scoped. Issue keys for services and agents, each bound to a role and revocable at any time.
- Self-service profiles. Members manage their own profile and credentials without a support ticket.
Who can do what, down to the type.
Roles and RBAC are first-class. Grant read and write per type, gate publishing, and keep an audit trail of every change, without paying for an enterprise tier.
- Per-type grants. Decide exactly which roles can read or write each content type.
- Publish gates. Separate who can draft from who can publish, with required-relation checks before anything goes live.
- Audit trail. Every create, update, and publish is attributed, so you always know who changed what.
Forms your team can build.
Model forms as content, add conditional logic, collect submissions, and localize every label. Wire submissions into flows to email, notify, or sync wherever you need.
- Build without code. Compose fields, sections, and conditional logic right in the admin.
- Submissions, handled. Capture, store, and export submissions, with validation you define once.
- Connected to flows. Trigger an automation on submit to email, tag, or push data to another system.
A media library that makes its own images.
Upload, organize, and localize assets, and generate or regenerate images with AI right where you use them. Index folders so agents can find the right asset through retrieval.
- Generate and regenerate. Describe an image and create it into your library, then regenerate until it is right.
- Organized and searchable. Folders keep assets tidy, and indexed folders feed RAG.
- A different asset per language. Localize media at the asset level and it applies everywhere that asset is used.
Every locale, first-class.
Opt in per field and Tangen stores each locale in real, typed columns, not a JSON blob. Blocks, media, and content all localize, with a public locales API for your front-end.
- Per-field opt-in. Localize only what needs it, stored in columns you can actually query.
- Blocks and media too. Per-locale overlays for blocks and per-asset media, not just plain text.
- One item, many languages. A single content item serves every locale through the API and SDK.
Retrieval your agents can cite.
Turn content, docs, and PDFs into a knowledge base with embeddings and pgvector. Agents answer from your data with citations, scoped to the audience you choose.
- Knowledge from your content. Index types, folders, and uploaded PDFs into a searchable base.
- Answers with citations. Responses point back to the source, so nothing is a black box.
- Audience-scoped retrieval. Retrieval respects who is asking, so private content stays private.
Your CMS, as tools for agents.
Tangen ships an MCP server so assistants read and write content through RBAC-scoped tools. Describe what you want and the assistant can build schema, draft content, or generate images.
- RBAC-scoped tools. Agents get exactly the access their role allows, and every write is attributed.
- Describe to build. Ask for a content type or a draft and the assistant creates it, ready for you to confirm.
- Per-type exposure. Choose which types agents can see, one switch at a time.
See the page as you write it.
Edit content and watch it render on your real site in real time. There is no publish-and-refresh loop, and no premium add-on to unlock it.
- Draft on the real site. Your front-end renders the draft exactly as it will look live.
- Included, not upsold. Live preview is part of the platform, not a paid tier.
- Bring any framework. Works with whatever front-end you put in front of it.
Everything else, included.
The infrastructure most platforms sell as add-ons. It ships in the box and stays out of your way until you switch it on.
The things others meter or paywall. Free here.
On many hosted platforms these arrive as higher tiers, add-ons, per-seat charges, or metered usage. In Tangen they are simply included, because you run it yourself.
Stop renting your content layer.
Hosted platforms bill per seat, per API call, and per premium feature, and it climbs with your team and your traffic. Tangen you run yourself, so you pay for the box it runs on. That is the whole bill.
$300 to $3,000+ / month
Once you add editor seats, API overages, roles, and multiple environments. It grows as you do.
Your infra only
Run it on a small VPS or the cloud you already pay for. No per-seat pricing, no metered calls, no feature paywalls. MIT licensed.
Ranges reflect common hosted headless-CMS pricing across seats, usage, and tiers; your actual costs vary. Tangen is free and open source, you provide the hosting.
Build on Tangen in the private beta.
Define a content type, push it, and query it from anywhere, with your own AI agent doing the rest. Request access and we will get you set up.