Open source
digital signage

Xibo, Anthias, Concerto, info-beamer, PiSignage, DigitalSignageOS and friends are free.

Keeping them alive for six months is what costs.

Honest map of the ecosystem with real numbers, no preaching, no bashing.

A Friday at 11:47 PM

Friday, 11:47 PM, the only technical contact for a three-cafe group in Manchester gets a phone call.

It's the closing manager. "The menu screen is dead. Black with red text." The contact opens the laptop from the sofa. SSH to the VPS. docker logs xibo-cms. A long stack trace pointing to a PHP update from two months ago that broke a dayparting module. He hadn't touched the instance since March. It worked until a maintenance cron from the hosting provider triggered a reboot and the container came up with a newer image.

It took him three hours to restore the database dump, downgrade the container to the previous version, and confirm that weekend playlists were running again. By 3:10 AM things were fine. Saturday morning the cafe opened with the menu on screen. Nobody noticed.

On Monday, in a chat with his manager, he ran the numbers. Third time in six months. He added up the hours: 28 hours of his time so far this year, not counting the stress. He ran another number: what it would cost to sign up to any cloud alternative, in pounds, against what his time was costing. First time the calculator came out in favour of cloud. He'd been telling himself Xibo was "free" for two years.

If you're reading this, you've probably had your own 11:47 PM Friday. Or not yet, but you sense it coming.

Let's run the whole calculation. No preaching. Starting by naming everything that's actually out there.

Find your section before going further

This post is long because it covers the full ecosystem. Pick your starting point:

• Researching, never installed anything open source → start with "The 2026 ecosystem map" and continue with "When open source IS the right answer".
• Already running something, it hurts → jump to "What hurts at month six" and "Migrating to SaaS without losing what works".
• Already decided to migrate → go straight to the "Migration audit checklist", and grab the TCO calculator linked at the end.

No open source CMS is bad in itself. Each one was built by somebody with a real problem who solved it well. What we're going to look at is which one fits which business reality. Why most European and UK SMBs end up moving to SaaS. And why some cases are right to stay self-hosted.

No clickbait. No "the big guys are lying to you". Just numbers and field experience.

The 2026 open source digital signage ecosystem map

A cleanup first: not everything sold as "open source alternative to Xibo" is actually open source. Let's separate.

The genuinely open source

1. Xibo

• Stack: PHP + MySQL/MariaDB + Docker. The CMS and players (Windows/Android/Linux/webOS) are AGPL from 2.x onwards.
• When it shines: medium installations (20-200 screens) with a sysadmin who already spoke PHP. Mature, has modules, native dayparting, multi-user with fine permissions.
• When it hurts: major upgrades break DB migrations if you skip more than two versions; the stats module hammers I/O; keeping the Android TV player working as Google tightens APK loading; the layout editor's learning curve is not for someone in marketing.
• Who chose well: museums, universities, multi-site retailers with in-house IT.
• Who regretted it: 2-5 screen SMBs without a dedicated sysadmin.

2. Anthias (formerly Screenly OSE / Screenly OSS)

• Stack: Python + SQLite + a single Chromium-based player. Designed for Raspberry Pi.
• When it shines: one screen, one Pi, one image or a slide. Setup in 20 minutes. If that's exactly what you need, it's hard to beat.
• When it hurts: one screen per Pi, no serious multi-screen management; active development is focused on the commercial Screenly product; the open source community gets patches sporadically.
• Heads up: in 2023 Screenly Inc. renamed the open source version to Anthias and separated it clearly from the commercial Screenly product. What you still see online as "Screenly OSS" today is Anthias. Repo: github.com/Screenly/Anthias.
• Who chose well: small gyms with one lobby screen, offices with a single notice display.
• Who regretted it: anyone expecting decent scheduling or multi-screen.

3. Concerto

• Stack: Ruby on Rails + PostgreSQL. Apache 2.0 licence.
• When it shines: born at Rensselaer Polytechnic Institute and the historical pick for American university campuses. Designed for places with lots of screens and cross-department collaboration.
• When it hurts: small community, slow roadmap, the frontend feels five years old, you have to be comfortable in Rails for patches.
• Who chose well: public universities, councils with technical departments, large NGOs.
• Who regretted it: shops and venues without a dev on payroll.

4. info-beamer

• Stack: Lua + a custom C runtime. Native for Raspberry Pi (a hosted commercial version also exists).
• When it shines: fastest renderer on this list. Heavy graphics, 4K video, complex transitions. Used by creative agencies and live events. Visual quality beats everything else here.
• When it hurts: layouts are Lua code. No proper visual editor in the free self-hosted version. CMS side has to be built separately or you use the paid hosted.
• Who chose well: creative agencies, events, art installations.
• Who regretted it: anyone expecting "upload a picture from my phone and it appears on the screen".

5. PiSignage Community Edition

• Stack: Node.js + Raspberry Pi + an Android client. Mixed model: open source Community edition and a paid hosted edition.
• When it shines: if your hardware is several Raspberry Pis and you want basic multi-screen orchestration, central management, and drag-and-drop playlists simpler than Xibo.
• When it hurts: Community edition lags behind the commercial; the roadmap clearly favours paying customers; open source support is "open an issue and pray".
• Who chose well: small chains with in-house technical staff.
• Who regretted it: venues without staff, because Community needs ongoing maintenance on the Pi side (SD card corruption, OS getting outdated, etc.).

6. DigitalSignageOS / DiSignageOS

• Stack: a stripped-down Linux distribution built solely to display content. More player/OS than CMS.
• When it shines: when you already have a CMS (or you're writing one) and want a hardened OS on the device, in kiosks, on near-unbreakable screens.
• When it hurts: you need the CMS on the side. Half a solution.
• Who chose well: integrators putting their own mini-PCs in client venues.
• Who regretted it: nobody coming for "a complete platform".

7. Now Playing System (NPS) and small SMIL projects

• Stack: small forks around the SMIL protocol (yes, SMIL is still alive in some niches).
• When it shines: very specific niches where you need raw SMIL (DOOH advertising, integrations with legacy retail players from very specific sectors).
• When it hurts: tiny community, scattered docs, high risk of abandoned project.
• Who chose well: organisations that were already speaking SMIL before 2015.
• Who regretted it: anyone starting today.

8. Pure DIY: Raspberry Pi + mpv/omxplayer + cron + Samba

• Stack: whatever you want. A Pi, a video player, a cron that rotates the file each hour, a shared folder where you drop next Monday's videos.
• When it shines: one screen, fixed content, total stack control, zero budget beyond hardware.
• When it hurts: any schedule change means SSH. Forget dayparting, multi-screen, templates, analytics.
• Who chose well: makers, hackerspaces, workshops with a single hours sign.
• Who regretted it: the first pub owner who thought it would be like Netflix.

Sold as "open source alternatives" but they aren't

Honest clarification because it helps:

• Yodeck is not open source. It's a commercial SaaS focused on Raspberry Pi with a free plan up to a certain screen count. Good people, good product, but calling it open source is incorrect.
• ScreenCloud, OptiSigns, NoviSign, Rise Vision, Mediasignage, OnSign TV, NetPresenter are commercial SaaS. Some have freemium, some have trials. None are open source.
• Screenly (without "OSS") is the current commercial version of the company that maintained Screenly OSS. The open source version became Anthias and lives separately. Don't mix them.

If your real question was "cheaper alternative to one of these SaaS", that's a different conversation. We cover it in dedicated posts per product (linked at the end).

The real cost of "free"

This is where most comparisons go wrong. They talk about licences and forget the rest.

Let's put numbers on three business sizes, with self-hosted Xibo on a small European VPS, three-year horizon. Assume external sysadmin at £30/h and owner at £25/h (your own time).

Case A: a cafe with 2 screens

Real annual cost:

Same case on Digisini Pro (3 screens, one spare for growth): Pro annual = €190/year × 3 years = £570. No hours. No downtime on you. Email support included.

Difference: ~£3,300 and 129 hours of people's time over three years, for a cafe with two screens.

Case B: small chain, 5 screens

Same case on Digisini Business (10 screens, 5 spare): Business annual = €590/year × 3 years = £1,770.

Difference: ~£5,500 and 234 hours of three people's time over three years.

Case C: mid-size network, 30 screens multi-site

Here the maths complicates in favour of self-hosted, if and only if you have in-house IT.

Same case on Digisini Network (50 screens, 20 spare): Network annual = £1,990/year × 3 years = £5,970.

Gross difference here is ~£6,780/year in favour of cloud, with a caveat: when you have 30 screens and a part-time paid sysadmin, that sysadmin probably has other duties beyond signage. The real number depends on whether that 15% is truly signage-only or you're double-counting.

When you hit 80-100+ screens with a full-time in-house sysadmin, self-hosted starts breaking even or winning on cost. But you're no longer an SMB. You're mid-size retail.

When open source IS the right answer

Cloud doesn't always win. Five profiles where staying (or moving into) open source self-hosted makes sense:

1. You have in-house IT and 30-50+ screens. Sysadmin cost is already paid for other things. Signage is just one more. And being able to modify the code is a real freedom, not a decorative one.

2. Non-negotiable data sovereignty. Banking, regulated defence, healthcare with data in specific countries, public sector with "data doesn't leave here" requirements. Cloud + compliance can be negotiated, but sometimes self-hosting is faster.

3. You need to extend the code. Your business depends on a module no SaaS will build for you. Exotic SAP integration, RFID readers in kiosks, syncing with the shopping centre's internal radio. Forking a project is reasonable then.

4. You're an agency or integrator selling the service. You sell clients a "your brand" solution where Xibo or info-beamer sits underneath. You take maintenance seriously because it's your business. There's a whole world of Xibo partners making a living from this.

5. You're a maker and learning is the reward. A Raspberry Pi and Anthias at home or in your workshop. It's fun. We're not here to judge fun.

If you recognise yourself in one of these five, this post might not be for you. Or it might, as recreational reading.

If you don't recognise yourself in any of them, keep reading. There's a very good chance the TCO is eating you alive without you noticing.

What hurts at month six (with Xibo and most self-hosted)

Anyone who's run Xibo in production for a year recognises this list. Anyone who's three months in will see it in another couple.

1. Updates break things. Xibo 4 changed so much under the hood that migrating from 3.x meant reading the changelog twice. info-beamer breaks APIs between major versions. Anthias rebranded and reorganised. Open source teams don't owe you eternal compatibility, and they don't give it. You end up stuck on an old version because you're scared to break Friday.

2. Dayparting is fragile. Works in Xibo, but when you have two time zones, five special events a year, and a manager who wants "evening menu out at 7:30 PM on Fridays and 8 PM on Saturdays", the editor groans. Anthias doesn't have it natively. info-beamer has it but you write Lua.

3. Support = Stack Overflow + Discord + luck. When something breaks at 9 PM, you open a browser and start searching. The community helps, yes. But it's not the same as having someone on the other side you pay and who cares your problem gets fixed.

4. Multi-site is an operational nightmare. Five venues, each with its VPN or static IP. Content push to the fleet. Resolving when each player has updated. Intermittent bugs in the furthest one. Any SaaS solves this in a week of use. Self-hosted forces you to engineer it.

5. No decent vertical templates. Xibo ships empty layouts. Want a cafe menu? You build it. A gym schedule? You build it. Multiply by all your venues and you understand why open source signage projects don't rank in SEO for "digital menu cafe" — they don't ship templates.

6. No allergen module, no queue management, no "Monday market at 11", no QR for bookings. Whatever your regulation or vertical needs, you build.

7. Player hardware. Android TV — Google is restricting external APKs. Windows — Windows updates kill kiosk mode sometimes. Linux — you manage it. Raspberry Pi — you play the corrupted SD card lottery every 18 months.

None of these six pains are unique to Xibo. They're inherent to the "I'll build it myself" model.

Migrating to SaaS without losing what works: audit checklist

If you decide to move to cloud, don't run out. Audit what you have first to avoid surprises:

Content inventory

• List of active playlists and where they appear
• Special events calendar (Christmas, Valentine's, football finals, fairs)
• Media folder organised by vertical and expiry date
• What content has to comply (allergens, 18+, legal notices)

Operational inventory

• How many active screens, where, what TV/player model
• Who changes content regularly (owner, manager, external marketing)
• How many hours that person spends monthly on this, plus yours
• How many technical incidents in the last 6 months
• What hardware you want to reuse (Smart TVs are your ally, Raspberry Pis maybe not)

Dependency inventory

• Anyone external (agency, freelancer, cousin) managing it
• Any custom script or module you'd need to rebuild
• Integrations with POS, booking software, order screens

Export from Xibo (or whatever you have)

• Download the media (everything in /var/www/.../library/)
• Export the playlists (Xibo lets you export layouts as .zip)
• Screenshot the current calendar before turning off, so you don't forget anything
• List of users and their access, if migration changes the permissions model

Audit done, signing up to a SaaS takes between 30 minutes and two afternoons. The long part was knowing what you had.

Where Digisini fits in this conversation

No bashing. Here's what Digisini is and who it's for. If the answer is "not me", I'll say it.

Real Free plan, no time limit.
1 screen, unlimited playlists, email support. Has a Digisini watermark in the corner. Easiest way to test the platform with a full month of your cafe before paying anything.

Start free →

14-day trial, no card, on every paid plan.
If you want to evaluate Pro or Business properly with all features, activate without a card. On day 15 you decide.

Pricing built for SMB, not enterprise.
Starter €9/month, Pro €19/month (3 screens, the bestseller), Business €59/month (10 screens), Network €199/month (50 screens). No commitment. VAT excluded. Compared to the real TCO of running Xibo we just saw, the numbers aren't really up for debate.

Human support in English (and Spanish if you prefer).
When something breaks, there's a real human on the other side. Not a chatbot asking you to fill a form before talking to you. The difference with "open an issue on GitHub" you notice on day one.

Vertical templates ready.
Cafe, gym, dental, driving school, pharmacy, pizzeria, hostel, B&B, campsite, accounting firm, law firm, garage, optician, funeral home, queue management for your counter: there's a template per sector with sector details (allergens where they apply, appointments, realistic dayparting, multi-language where relevant). 37 verticals built specifically, so you don't reinvent the layout each time.

Works on the TV you already have.
Smart TV with a browser, Chromecast (~£25) or Fire Stick (~£30). Zero proprietary hardware. Zero installers.

No commitment. Cancel anytime.
If three months in you discover self-hosting was actually your thing after all, you go back. No drama.

Migration assistance — honest version.
We give you step-by-step guidance to export from Xibo (or whatever you have) and rebuild the same structure in Digisini, with 1:1 human support during the migration by email or WhatsApp. What we don't say is "send us your VPS and we'll migrate it for you" because today we can't guarantee that service to everyone who'd ask. If we can ever automate it, we'll tell active customers first.

Two honest cases side by side

Case 1: when Xibo wins

A Manchester digital agency with four big corporate clients (a bank, a multi-site retailer, a private university, an events firm). 280 screens across the four. Three devs on payroll. Five years on Xibo, three custom modules written in-house (one for the bank's queue system integration, one for the university's academic calendar, one to sync with the retailer's data warehouse).

For them, Xibo is the right tool. Code control lets them sell maintenance contracts, the custom modules are their differentiator vs other agencies, infra costs are passed through to the client.

If they ask us, we say: keep Xibo. You're not our target. We'll take your call if one client ever doesn't fit your model, but the bulk of your business is fine where it is.

Case 2: when Digisini wins

A small sandwich shop chain in the north of England. Twelve venues, one screen behind each counter for the daily menu and allergens, and one larger screen in the seating area for promos and events. The owner had a cousin who knew IT and installed Xibo on a VPS two years ago. The cousin moved to Berlin seven months ago. Since then, every update is a drama because nobody wants to touch it, allergens are updated by hand venue by venue via USB stick, and the last time a menu went stale on social media generated a viral negative comment that cost them a morning of community management.

For them, the day they move to Pro (3 screens) or Business (10 screens, covering 8 of the 12 until they grow to Network), everything changes: the owner changes the daily menu from his phone each evening for next day, allergens propagate to 12 screens at once, the Valentine's event sign is scheduled by his niece studying marketing from Leeds so it appears exactly Feb 12-14, and nobody in the venue touches a USB stick again.

Not marketing. Real case (with changed names) from 2025.

Your next step (no pressure)

Made it this far, one of three things fits you:

Want data before you talk

Grab the TCO Calculator: Digital Signage Open Source vs SaaS. Excel with your real numbers (screens, salaries, hours), gives you three-year cost in each model. Sheet for Xibo, for self-hosted Anthias, for self-hosted info-beamer, for SaaS. Email not required, but if you leave it we'll let you know when we update the market figures (VPS, salaries, hardware) so your calculation stays current.

Download calculator + PDF guide →

Free, immediate. Email required, both files arrive in seconds.

Rather try first

Sign up for the Free plan. One screen, watermark, 5 minutes setup. If you like it, you upgrade to Starter (€9/month) anytime and the watermark disappears.

Start free →

Want to talk to a person

Email [email protected], a human responds in English (or Spanish, your call) within European working hours. We'll tell you what fits your case, and if the honest answer is "stick with Xibo", we say it.

Last updated: 2026-05-30. This post is maintained. If something here goes out of date, write to [email protected] and we'll fix it. Honour of the trade.