bewCloud

Self-hosted personal cloud, honestly reviewed. No marketing fluff, just what you get when you run it yourself.

TL;DR

• What it is: Open-source (AGPL-3.0) personal cloud platform — files, notes, news reader, photos, expenses, calendar, and contacts — built with TypeScript and Deno as a deliberate reaction to Nextcloud’s bloat [README][website].
• Who it’s for: Solo founders, individuals, and small households who want to replace Google Drive, Google Calendar, and Google Photos with something self-hosted, lightweight, and maintainable by one person [README].
• Cost savings: Google One runs $2.99–$9.99/mo for storage tiers. iCloud+ is $0.99–$9.99/mo. bewCloud self-hosted runs on a $5–10/mo VPS with no per-user pricing and no per-GB fee [README].
• Key strength: Dramatically smaller resource footprint than Nextcloud or ownCloud at idle. Built by someone who ran Nextcloud for years, got frustrated with its PHP complexity and bloat, and rebuilt only the parts they actually used [website].
• Key weakness: This is a one-developer project (Bruno Bernardino + community contributors) with 1,128 GitHub stars. It is not enterprise software. It does not have SSO, audit logs, LDAP, or a plugin ecosystem. The AGPL-3.0 license means if you embed it in a commercial product, you must open-source that product [README][1].

---

What is bewCloud

bewCloud is a personal cloud application covering the core things people actually use a cloud for: file storage and sync, a news reader, notes, photos, an expenses tracker with envelope budgeting, CalDAV-based calendar, and CardDAV-based contacts. The README describes it as “a simpler alternative to Nextcloud and ownCloud, built with TypeScript and Deno” and that description is completely accurate — no more, no less [README].

The project was started in early 2024 by Bruno Bernardino, a developer who had been running Nextcloud for years and found himself accumulating a list of annoyances he never fixed because Nextcloud’s PHP codebase was too complex to touch. Rather than fork it, he started fresh. By his own account, “it felt like such a daunting task that I never even started it” — until he decided he was OK if this was a never-ending project that just worked for him and his family [website]. That origin matters for setting expectations: bewCloud is purpose-built to solve a real person’s real problem, not to capture enterprise customers or VC funding.

The name is an anagram: bew is an anagram of web, the same way deno is an anagram of node [website]. The technology stack reflects those choices — TypeScript throughout, Deno as the runtime instead of Node.js, PostgreSQL for storage, and Docker Compose for deployment. There’s a separate desktop sync client (bewcloud-desktop) and a mobile app (bewcloud-mobile), both maintained in their own repositories [README].

The project has received a grant from the NLnet Foundation, which funds internet freedom and privacy infrastructure across Europe, and is sponsored by Uruky (a private search engine) [README]. That’s a meaningful signal — NLnet doesn’t fund projects at random, and having that backing gives the project some institutional credibility beyond a solo developer’s side project.

As of this review it sits at 1,128 GitHub stars [merged profile].

---

Why people choose it

The bewCloud positioning is simple enough that the “why” is almost self-evident: people are running Nextcloud, it’s slow, it consumes a lot of RAM even when idle, and they don’t use 90% of its features anyway.

The website’s own pitch says it plainly: Nextcloud and ownCloud have large resource footprints “partly because they have a lot of legacy to support, and they’re very customizable and extensible” [website]. That is accurate. A default Nextcloud install with a few apps enabled idles at 300–500MB RAM. For someone on a $5 Hetzner instance with 2GB RAM, that’s a meaningful chunk. bewCloud was built to be lean by default.

There are no deep third-party reviews with user benchmarks or detailed performance comparisons available for this review [1]. What AlternativeTo shows is that bewCloud has been added as an alternative to Nextcloud, ownCloud, Tuta Drive, and other cloud storage options by users who are clearly looking for the same thing: a self-hosted personal cloud that doesn’t require a dedicated server just to idle [1]. The feature list on AlternativeTo — WebDAV support, file sync, two-factor authentication, dark mode, selective synchronization, shared folders — matches what the README actually delivers [1][README].

The honest picture: bewCloud isn’t chosen because it’s the best at anything specific. It’s chosen because it’s the simplest thing that covers the bases for a person or small household exiting Google or Apple cloud services. It is not the tool you pick when you need Nextcloud’s depth. It’s the tool you pick when Nextcloud’s depth is precisely what you’re trying to escape.

---

Features

Based on the README and website:

File management:

• File upload, storage, and browsing via web UI [README]
• Desktop sync client (separate bewcloud-desktop repo) [README]
• Mobile app (separate bewcloud-mobile repo) [README]
• WebDAV support for third-party clients [1]
• Shared folders [1]
• Selective synchronization [1]

Personal productivity:

• News reader — RSS-style feed aggregation [website]
• Notes [website]
• Photos gallery [website]
• Expenses with envelope budgeting [website screenshots]
• Calendar via CalDAV [website screenshots][README]
• Contacts via CardDAV [website screenshots][README]

Auth and access:

• Two-factor authentication [1]
• Admin account created on first signup even with signups disabled [README]
• Sign-up toggle (config.auth.allowSignups) [README]

Technical:

• Docker Compose deployment [README]
• Helm chart available [merged profile]
• PostgreSQL backend [merged profile]
• Radicale (Python CalDAV/CardDAV server) bundled for calendar and contacts [README]
• TypeScript + Deno stack [README]
• Dark mode [1]
• Multi-language UI [website]

What it doesn’t have:

• SSO / SAML / LDAP — not mentioned anywhere in the available documentation
• Plugin/app marketplace — bewCloud does not have a Nextcloud-style app ecosystem
• Collaborative document editing — no equivalent of Nextcloud Office or OnlyOffice
• Video calls or messaging
• Team permissions or role-based access control at any meaningful granularity

The scope is intentionally narrow. This is a personal cloud, not a team productivity platform [website].

---

Pricing: SaaS vs self-hosted math

bewCloud managed cloud:

• The website sells a 1-year managed (dedicated) cloud instance via Mollie payment links [README][website]
• Specific annual pricing is not published on the homepage — it requires clicking through to the payment link
• Data not publicly available for exact figures at time of writing

Self-hosted:

• Software license: $0 (AGPL-3.0) [README]
• VPS: $5–10/mo on Hetzner, Contabo, or DigitalOcean
• PostgreSQL: bundled in default docker-compose [README]
• Domain + TLS: minimal cost if you already have infrastructure

Google One for comparison:

• 100 GB: $2.99/mo
• 2 TB: $9.99/mo
• Covers Drive, Photos, Gmail — but you’re trusting Google with everything

iCloud+ for comparison:

• 50 GB: $0.99/mo
• 200 GB: $2.99/mo
• 2 TB: $9.99/mo

Concrete math: A family of four sharing a 2TB Google One plan pays $9.99/mo = $120/year. Self-hosting bewCloud on a $6 Hetzner VPS with 2TB attached storage comes to $72–120/year depending on storage volume — roughly break-even on the Google plan, but you own the infrastructure and there’s no per-user charge. The savings case is stronger if you’re currently paying for multiple separate services (cloud storage + notes app + calendar app), each of which bewCloud covers in a single install.

The more honest framing: bewCloud’s cost advantage over consumer cloud services is modest in absolute terms. The real reason people self-host it is control and privacy, not purely the money [website][1].

---

Deployment reality check

The install path documented in the README is Docker Compose — download four files, run two commands, done [README]. There’s no one-line installer, no web-based setup wizard, no managed Kubernetes that someone else runs. You get a docker-compose.yml, a .env file, and a bewcloud.config.ts.

What you actually need:

• A Linux VPS (RAM requirements not explicitly documented — given the Deno/TypeScript stack and PostgreSQL, 1GB minimum, 2GB recommended is a reasonable estimate)
• Docker and docker-compose
• A domain name and reverse proxy (Caddy or nginx) for HTTPS
• Basic shell comfort to run docker compose up and make migrate-db

The CalDAV/CardDAV piece is a separate service. Radicale — a Python CalDAV/CardDAV server — is bundled in the docker-compose setup, but you need to copy its config file manually. The README is clear about this: “necessary only if you’re using CalDav/CardDav/Contacts” [README]. Skip the step and you won’t have calendar or contacts sync.

Known permission gotcha: If you run into file permission errors, the fix is sudo chown -R 1993:1993 data-files — the 1993 UID comes from Deno’s Docker image and could change in future upstream releases [README]. This is a minor but real operational wrinkle worth knowing before you’re debugging at 11pm.

Database migrations must be run manually on first install and after any data updates: docker compose run --rm website bash -c "cd /app && make migrate-db" [README]. There is no auto-migration on startup.

Realistic time estimate for a technical user: 30–60 minutes to a working instance. For a non-technical founder following a step-by-step guide: 2–4 hours including domain and TLS setup. If you’ve never run Docker on a Linux server, plan for a full afternoon.

No major community horror stories about installation failures surfaced in the available sources [1], which is itself a signal that the install process isn’t catastrophic.

---

Pros and Cons

Pros

• Genuinely lightweight. Built from scratch to be lean, unlike Nextcloud which carries years of PHP legacy and plugin infrastructure [website].
• Coherent scope. Files, notes, news, photos, expenses, calendar, contacts — that’s it. No 500-app marketplace where 400 apps are abandoned. What’s there, works [README][website].
• TypeScript + Deno stack. Modern, readable codebase that a JS/TS developer can actually read and modify without a PHP archaeology expedition [README].
• WebDAV support. Means your existing file manager or sync client (Cyberduck, rclone, Mountain Duck) connects without needing the native app [1].
• Mobile and desktop apps exist. Both are separate repos but maintained by the same developer [README]. Most lightweight Nextcloud alternatives don’t have this.
• NLnet-funded. Institutional backing from a respected European internet freedom foundation adds credibility to the project’s longevity [README].
• Managed cloud option. If you want bewCloud without running your own server, the developer sells a 1-year managed instance [README][website].
• Built by someone who actually uses it. Bruno explicitly said he uses it himself and for his family [website]. That’s the best quality signal available for a solo-developer project.

Cons

• AGPL-3.0, not MIT. If you embed bewCloud in a commercial product or SaaS, you must open-source that product under AGPL. This is a real legal constraint that MIT alternatives like OxiCloud don’t have [README][3].
• One primary developer. Bruno Bernardino is the main contributor [website]. There are community contributors and NLnet backing, but the bus factor is high. If he stops, the project’s future is uncertain.
• 1,128 GitHub stars — small ecosystem [merged profile]. Compare to Nextcloud’s 30K+. Fewer community integrations, fewer tutorials, fewer people who’ve hit your exact error.
• No SSO / LDAP / SAML. Not mentioned in documentation. Not suitable for any organization that manages users centrally.
• No collaborative editing. No OnlyOffice, no Collabora. Files are stored and synced, not edited collaboratively in-browser.
• No plugin system. The feature set is fixed to what the developer built. You can’t extend it with apps the way you can Nextcloud.
• CalRADICALE as a separate process adds operational complexity — another service to manage, another config to maintain [README].
• Pricing for managed cloud is not transparent. You have to click through a payment link to find out what you’d pay. That friction is unusual and slightly suspicious for a modern product [README][website].
• No third-party reviews with hands-on testing. The sources available for this review are AlternativeTo listings, not independent in-depth tests [1][2][3][4][5]. Any review of bewCloud right now (including this one) is working mostly from primary sources.

---

Who should use this / who shouldn’t

Use bewCloud if:

• You’re a solo founder, freelancer, or individual who wants to escape Google Drive and Google Calendar with a single self-hosted install.
• You have basic Docker comfort or someone who can deploy it for you once.
• You want a personal cloud covering files, notes, photos, and calendar without the operational overhead of Nextcloud.
• You don’t need team governance, SSO, or a plugin marketplace.
• You’re fine with a small project backed by one developer, and you’d rather run lean software than feature-rich bloatware.

Skip it (try Nextcloud instead) if:

• You need SSO, LDAP, SAML, audit logs, or fine-grained user management.
• You need collaborative document editing in-browser.
• You need a plugin ecosystem (Deck, Talk, Mail, integrations with Jira or Slack).
• You’re deploying for a team of more than 2–3 people.
• You need the confidence that comes with a large community and years of production hardening.

Skip it (try OxiCloud instead) if:

• You want a MIT-licensed file storage alternative with no AGPL encumbrance and are comfortable with a Rust-based stack [3].

Skip it (try Hoodik instead) if:

• Your primary concern is end-to-end encryption — Hoodik encrypts on-device via WebAssembly before anything reaches the server, which bewCloud does not do [4].

Skip it (stay on Google/iCloud) if:

• You’re not comfortable with the command line and don’t have anyone to help with deployment.
• You have fewer than 5 files you care about and the free tier of any consumer cloud covers you.
• Your threat model doesn’t actually require self-hosted infrastructure — most people’s doesn’t.

---

Alternatives worth considering

• Nextcloud — the incumbent. Full-featured, massive community, huge plugin ecosystem, available managed hosting from multiple providers. Heavier resource footprint, PHP codebase, significant complexity. The thing bewCloud was built as a reaction to [website].
• ownCloud — similar age and legacy to Nextcloud (it was forked from ownCloud). Recently acquired by Kiteworks. Enterprise-focused, not the right choice for personal use anymore [2].
• OxiCloud — Rust-based, MIT-licensed, lightweight file storage focused on speed and simplicity. Fewer built-in features than bewCloud (no notes, no expenses, no CalDAV) but a cleaner license for commercial use and 2,946 GitHub stars vs bewCloud’s 1,128 [3].
• Hoodik — end-to-end encrypted file storage with WebAssembly-based client-side encryption, mobile apps, and a built-in encrypted note editor. 1,106 GitHub stars, more security-focused than bewCloud, less feature-breadth [4].
• Seafile — more mature file sync platform with good desktop clients, better suited for multi-user environments than bewCloud.
• Syncthing — if all you need is file sync without a web UI or server, Syncthing is simpler, faster, and has a huge community. No CalDAV, no notes, just sync.

For a non-technical founder trying to replace Google’s suite with a single install, the realistic shortlist is bewCloud vs Nextcloud. Pick bewCloud if you want simplicity and are comfortable with a smaller community. Pick Nextcloud if you need depth, plugins, and production-grade multi-user support.

---

Bottom line

bewCloud is an honest project solving a real problem: Nextcloud is too heavy for personal use, and the founder proved it by building a replacement that actually runs light. The scope is correct — files, notes, news, photos, expenses, calendar, contacts — and it covers the Google Suite basics without a PHP monolith underneath. The trade-offs are equally honest: one primary developer, 1,128 stars, no SSO, no plugins, an AGPL license that limits commercial use. For a solo founder or household replacing Google services on a $6 VPS, the math and the architecture both work. For a team of ten needing centralized user management and collaborative editing, this isn’t the tool yet and may never be.

If the deployment is the blocker, that’s exactly what unsubbed.co’s parent studio upready.dev handles for clients. One-time setup, you own the server, done.

---

Sources

1. AlternativeTo — bewCloud: Self-hosted open-source cloud storage with WebDAV support (5 likes, 56 alternatives listed). https://alternativeto.net/software/bewcloud/about/
2. AlternativeTo — ownCloud: Self-hosted file sync and sharing (868 likes, 8 reviews). https://alternativeto.net/software/owncloud/about/
3. AlternativeTo — OxiCloud: Lightweight file storage with Rust backend (11 likes, 65 alternatives listed, 2,946 GitHub stars). https://alternativeto.net/software/oxicloud/about/
4. AlternativeTo — Hoodik: Encrypted self-hosted cloud storage with E2E encryption (28 likes, 109 alternatives listed, 1,106 GitHub stars). https://alternativeto.net/software/hoodik/about/
5. AlternativeTo — Tracim: Team collaboration tool (18 likes, 62 alternatives listed). https://alternativeto.net/software/tracim/about/

Primary sources:

• GitHub repository and README: https://github.com/bewcloud/bewcloud (1,128 stars, AGPL-3.0 license)
• Official website: https://bewcloud.com
• Desktop sync client: https://github.com/bewcloud/bewcloud-desktop
• Mobile app: https://github.com/bewcloud/bewcloud-mobile
• NLnet Foundation grant page: https://nlnet.nl/project/bewCloud/