Every version. Every file.
Always recoverable.

AeorDB is a content-addressed file database with built-in versioning. Store any file, track every change, roll back anything — from a single binary with zero dependencies.

Get Started View Documentation

Sound familiar?

presentation_final_v3_ACTUALLY_FINAL_revised.psd
A shared drive where nobody knows which version is current
Colleagues overwriting each other's work with no way to undo
A project folder with 47 copies of "the latest" file
Large assets that outgrow every tool you've tried

Your files deserve history.

Built For Your Workflow

🎨

Creative Studios

Version every asset from concept to delivery. Compare revisions, fork for experimentation, never lose a comp.

⚖️

Legal & Compliance

Immutable audit trail for every document revision. Know exactly who changed what, and when.

📣

Marketing Teams

Campaign assets across dozens of stakeholders. Branch for regional variants, merge when approved.

🎧

Music & Audio Production

Track every mix, every stem, every master. Roll back to any session instantly.

💚

Nonprofits & NGOs

Free for organizations under $2M. Manage grant documents, donor records, and program files with full version history — no licensing costs.

🏥

Healthcare & Research

Maintain audit trails for clinical data, research papers, and imaging files. Every version preserved, every change traceable.

🏗️

Architecture & Engineering

CAD files, blueprints, and specs across revision cycles. Snapshot before each client review, restore any version on demand.

🎓

Education & E-Learning

Course materials, lecture recordings, and student submissions. Version curricula across semesters, track every update.

Under The Hood

Powerful features for developers and teams.

Content-Addressed Storage

Every file identified by its BLAKE3 hash. Built-in deduplication, integrity verification on every read, and automatic corruption detection.

Global Search

Search across every file in the database with a single API call. Default indexes on filename, size, hash, and content type are created automatically on first boot.

Multi-Node Replication

Any node can write. Nodes sync by exchanging content hashes and merging automatically. Work offline, sync when connected. No leader, no quorum.

Git-Like Versioning

Snapshots, forks, diff, patch, rename/move, export/import. Restore any file from any version with a single API call. Full history without the large-file headaches.

Conflict Resolution

Concurrent edits are detected, preserved, and resolved — never silently lost. Modifications always beat deletions. Your work is safe.

Native File Parsing

Built-in parsers for 25+ file formats — PDF, Office, images, audio, video, HTML, and more. Automatic metadata extraction during indexing, no plugins required.

Access Control

Scoped API keys with path-based permissions and glob patterns. Tag-based user groups for team-level access control. Read-only, project-scoped, and time-limited keys.

Single Binary, Zero Dependencies

One executable, one database file, native TLS/HTTPS. Configure via TOML file or CLI flags. Like SQLite, but for versioned files with replication. Crash recovery and health monitoring built in.

Zero-Downtime Growth

The KV index grows online — 64 KB to 128 MB in geometric stages — without a restart, rebuild, or service interruption. Reads and writes continue while the database expands itself in place.

Live User Notifications

A JWT-gated per-user SSE channel delivers events addressed to a specific user — file shares, group invitations, mentions. The server filters by recipient before delivery, so users only see their own notifications.

Symlinks & Selective Sync

Soft symlinks with transparent resolution and cycle detection. Sync only the folders you need — each client or node chooses its own paths.

Real-Time Monitoring

Live metrics streamed over SSE, instant O(1) stats, rolling throughput rates, and a Prometheus-compatible endpoint. Know exactly what your database is doing, always.

Simple, Fair Licensing

Free for most teams. Commercial license for larger organizations.

🏠

Community

Free — forever

For organizations under $2M/year in gross revenue. Full product, all features, self-hosted. No feature gates, no time limits.

💼

Commercial

$500/mo per production instance

For organizations over $2M/year. Includes license compliance, email support, and priority bug fixes.

🏙

Enterprise

Contact us

Custom SLAs, dedicated support, consulting, and training. Tailored to your organization's needs.

Contact Sales

Frequently Asked Questions

Is AeorDB open source?

AeorDB is source-available under the Business Source License 1.1 (BSL). The full source code is public on GitHub. Every version converts to Apache 2.0 (fully open source) four years after release.

Can I use it for free?

Yes. If your organization's gross annual revenue is under $2M, you can use AeorDB in production for free — all features, no limits. Individuals, nonprofits, and startups are covered.

What requires a commercial license?

Three things: organizations over $2M/year in revenue, anyone offering AeorDB as a managed service to third parties, or anyone building a competing product derived from AeorDB's source code.

Can I modify the source code?

Absolutely. You can modify, fork, and create derivative works. Non-production use (development, testing, staging) is always free regardless of company size.

What happens after 4 years?

Each version converts to Apache 2.0 four years after release. At that point, all restrictions are removed for that version. Newer versions remain under BSL until their own clock expires.

Can I self-host for my company?

Yes. Self-hosting is the primary use case. If your organization is under the revenue threshold, it's completely free. Over the threshold, a commercial license covers all your instances.

Get Started

Two pieces. Pick what you need — or grab both.

AeorDB Server

The database engine itself. One binary, no dependencies. Run it as a standalone service, embed it in your stack, or both. This is what you install on the machine that stores and serves your files.

Server — Linux (x86_64) Server — Windows (x86_64) Server — macOS (Apple Silicon)

Manual updates — check back here for new releases. Intel Macs and other platforms: build from source.

After download — 60-second walkthrough

Start the server. Make it executable and run. The root API key prints once on first start — save it.
```
chmod +x aeordb-linux-x86_64
./aeordb-linux-x86_64 start --database mydata.aeordb
```

Store + retrieve files over HTTP. Any HTTP client works — no SDK required.

curl -X PUT http://localhost:6830/files/report.pdf \
  -H "Authorization: Bearer YOUR_TOKEN" \
  --data-binary @report.pdf

↪ Or use the AeorDB Client for a GUI.

Open the built-in dashboard. Live metrics, file counts, throughput — in your browser.
```
http://localhost:6830/system/portal
```
Go deeper. Snapshots, forks, replication, TLS, plugins — full documentation.

AeorDB Client

Desktop app that connects to an AeorDB server. Browse files, run queries, manage snapshots and forks from a GUI. Optional — everything the client does is also available over HTTP, but the GUI is faster for day-to-day work.

Client — Linux (x86_64) Client — Windows (x86_64) Client — macOS (Apple Silicon)

Auto-updates via a signed release channel — the client only installs binaries whose ed25519 signature verifies against its bundled trust store.