Fredy

Open-source real estate scraping, honestly reviewed. No marketing fluff, just what you get when you run your own listing alert bot.

TL;DR

• What it is: Self-hosted scraper that monitors five major German real estate platforms simultaneously and fires instant notifications when new listings match your search [1][2].
• Who it’s for: Anyone actively apartment or house hunting in Germany — especially in high-demand cities like Berlin, Munich, Hamburg, or Frankfurt where listings disappear within hours.
• Cost savings: Fredy itself is free (Apache-2.0). A VPS to run it costs $5–10/month. Manual platform premium tiers — ImmoScout24 Premium runs €39.95/month — add up fast for a multi-month search [2].
• Key strength: Covers ImmoScout24, Immowelt, Immonet, eBay Kleinanzeigen, and WG-Gesucht in a single tool, with deduplication so you don’t see the same listing twice across platforms [1].
• Key weakness: Germany-only by design. Scraping is inherently fragile — platform changes can break providers without notice. The ImmoScout24 integration relies on a reverse-engineered mobile API, which sits in a legal gray zone [1].

---

What is Fredy

Fredy is a self-hosted real estate notification bot built specifically for the German housing market. The name is an acronym: [F]ind [R]eal [E]state [D]amn Eas[y] [1]. It runs continuously in the background, polls supported platforms at configurable intervals, and pushes new listings to you the moment they appear — via Slack, Telegram, email, Discord, or ntfy — before you’d ever open a browser to check manually.

The project lives at GitHub under the MIT-compatible Apache-2.0 license, has 562 stars, and is maintained by a single developer (orangecoding) backed by the JetBrains Open Source Support Program [1][2]. That’s not a funded startup — it’s one developer scratching a real itch and keeping it alive. The project has enough momentum that it made it into the Unraid community app store, which is a reasonable proxy for “it works reliably enough that a hobbyist homelab community adopted it” [1].

The core problem Fredy solves is specific and acute: the German rental market, particularly in major cities, is one of the most competitive in Europe. A 2-bedroom apartment in Munich or Berlin can attract 200+ applications within 24 hours of posting. Speed is not a courtesy — it’s a prerequisite for even getting a viewing appointment. Checking five different platforms manually every hour is not a sustainable strategy during a months-long search. Fredy automates that watch [2].

---

Why People Choose It

There are no independent third-party reviews available for Fredy at time of writing — it’s too niche a tool to have attracted technology press coverage. What’s clear from the README and website is that it was built because existing options are bad in specific ways [1][2]:

The manual checking problem. Real estate platforms in Germany don’t have a unified feed. ImmoScout24 is the dominant aggregator but doesn’t list everything. WG-Gesucht is specifically for shared apartments (Wohngemeinschaft) and has its own user base. Immonet and Immowelt serve distinct regional audiences. eBay Kleinanzeigen picks up private landlords who don’t want to pay listing fees. To see the full market, you check five sites. Fredy checks them all for you, continuously [1].

The notification problem. Platform native alerts are slow and coarse. ImmoScout24’s own notification emails arrive with noticeable delay and lump results in batches. Fredy fires instantly — or as fast as your configured interval — and delivers to the channel you’re already watching (Telegram is the obvious choice for most users hunting apartments) [1].

The duplication problem. Many landlords post the same property on multiple platforms simultaneously. Without deduplication, you’d get alerted three times for the same apartment. Fredy tracks what it’s already shown you across all providers [1].

The cost problem. Premium tiers on German real estate platforms are designed to extract money from desperate renters. Fredy replaces that with a $5 VPS and an afternoon of setup [2].

---

Features

Based on the README and official documentation [1][2]:

Platform coverage:

• ImmoScout24 (via reverse-engineered mobile API)
• Immowelt
• Immonet
• eBay Kleinanzeigen
• WG-Gesucht

Notification channels:

• Slack
• Telegram
• Email (SendGrid and Mailjet supported)
• Discord
• ntfy

Core engine:

• Web UI for managing search jobs and configuration — no config-file-only setup [1]
• Customizable search intervals per job
• Cross-platform deduplication — once a listing is shown, it won’t be shown again [1]
• Persistent storage of found listings (SQLite-based) so results survive restarts

Infrastructure:

• Docker (single-command deployment) [1]
• Node.js 22+ for manual installation [1]
• Unraid community store [1]
• REST API and webhooks (listed as canonical features) [merged profile]
• Live demo available at fredy-demo.orange-coding.net [1]

What it doesn’t do:

• No AI-based relevance filtering (it matches by the filters you set on the source platform, not by semantic matching)
• No map-based filtering inside Fredy itself — you rely on the platform’s own search parameters
• No email-to-landlord automation — you still click the listing and respond yourself
• No support for Austrian or Swiss platforms (atWonen, Homegate, etc.)

---

Pricing: SaaS vs Self-Hosted Math

Fredy has no SaaS tier. It’s self-hosted software, period. The cost math compares it against alternatives:

Fredy self-hosted:

• Software: $0 (Apache-2.0)
• VPS to run it: $5–10/month on Hetzner, Contabo, or DigitalOcean
• Your time to set up: 30–90 minutes

What you’re replacing:

ImmoScout24 Premium Mitgliedschaft: approximately €39.95/month (at time of writing). This buys you priority contact with landlords and search agents — not a multi-platform monitoring bot. It doesn’t watch Immowelt or WG-Gesucht for you.

Immowelt Plus / Immonet subscriptions: each platform has its own premium tier ranging €10–30/month, none of which cross-monitor competitors.

Manual daily checking: not a monetary cost, but 30–60 minutes per day over a 3–6 month apartment search in a competitive city is 45–90 hours of your time. At any reasonable freelance rate, that’s a significant implicit cost.

Concrete scenario: You’re job-relocating to Munich and have 8 weeks to find an apartment before your start date. You need to watch all five major platforms around the clock. Alternative cost: one ImmoScout24 Premium subscription (€40/month × 2 months = €80) plus daily manual checking of the other four platforms. Fredy cost: $10 VPS for 2 months = $20 plus setup time. Savings: ~$60 in subscriptions plus however you value two months of daily manual checking [2].

The math isn’t dramatic in absolute euros — the real value is the time and the coverage. No premium tier gets you real-time alerts across all five platforms simultaneously.

---

Deployment Reality Check

The Docker path is the one you should take [1]:

docker run -d --name fredy \
  -v fredy_conf:/conf \
  -v fredy_db:/db \
  -p 9998:9998 \
  ghcr.io/orangecoding/fredy:master

This gets you a running instance on port 9998. Default credentials are admin / admin — change these immediately [1]. You’ll need to mount a config.json before Fredy starts, and the README points you to the template in the repo [1].

What you actually need:

• Any Linux VPS with Docker installed (512MB RAM is likely sufficient)
• A domain and reverse proxy (Caddy or nginx) if you want HTTPS
• An account with SendGrid or Mailjet if you’re using email notifications

What can go sideways:

The biggest reliability risk is scraping fragility. Real estate platforms actively change their HTML structure, API responses, and rate limiting. The ImmoScout24 provider specifically uses a reverse-engineered mobile API [1] — that’s not a stable public contract. When ImmoScout24 updates their app, the integration can break. This is not hypothetical; it’s the chronic maintenance burden of any scraping project. You should expect occasional periods where one or more providers stop working until the maintainer pushes a fix.

With a single maintainer and 562 stars, response time to provider breakages depends entirely on that person’s availability. This isn’t a criticism — it’s just the honest reality of a small open-source project. Check the GitHub issue tracker before committing to it for a time-sensitive search.

The Apache-2.0 license covers your use of the software. It does not cover the legality of scraping the platforms. German courts have had mixed rulings on automated scraping, and ImmoScout24’s terms of service prohibit automated access. This is an area where Fredy’s “reverse-engineered mobile API” note [1] is quietly significant. For personal use during an apartment search, enforcement risk is essentially zero. This would be a different conversation if you were building a business on top of it.

Realistic time estimate for a technical user: 30–60 minutes to a working instance. For someone following the README carefully without prior Docker experience: 2–3 hours including domain and notification setup.

---

Pros and Cons

Pros

• Apache-2.0 licensed. Genuinely free to run, modify, and self-host with no commercial strings [1]. This matters if you want to run it for family members or a group apartment search without worrying about license terms.
• Five platforms, one tool. The combination of ImmoScout24 + Immowelt + Immonet + eBay Kleinanzeigen + WG-Gesucht covers the meaningful German market. No alternative does this in a single self-hosted package [1].
• Six notification channels. Telegram, Slack, Discord, ntfy, email (SendGrid/Mailjet) [1]. If you’re reachable somewhere, Fredy can reach you there.
• Cross-platform deduplication. The detail that sounds minor but matters daily — you don’t get the same listing pushed five times when a landlord multi-lists [1].
• Web UI for configuration. Not a pure config-file tool. You manage search jobs through a browser interface [1][2], which meaningfully lowers the bar for less technical users.
• JetBrains-backed, demo available. Signals that the project has enough legitimacy to attract institutional support. A live demo at fredy-demo.orange-coding.net lets you evaluate it before committing [1][2].
• Unraid community store. Indicates broader community adoption beyond the GitHub audience [1].

Cons

• Germany-only, by design. The supported platforms are all German. If you’re searching in Austria, Switzerland, or anywhere else, this tool doesn’t help [1][2].
• Scraping is fragile. Provider integrations break when platforms change their structure. Single-maintainer response time is variable [1].
• ImmoScout24 reverse-engineered API is a legal gray area. Works today; terms of service doesn’t permit it; enforcement for personal use is unlikely but nonzero [1].
• No relevance filtering inside Fredy. You rely on the source platform’s filter parameters. Fredy shows you everything matching your platform-side criteria — there’s no secondary “only show me listings mentioning ‘pets allowed’” layer.
• 562 stars. Small community. Documentation is adequate but not extensive. Community support on edge cases will be limited.
• Single maintainer. Not a funded project. Longevity and issue response time depend on one person’s bandwidth.
• No cross-notification deduplication tracking. If the same listing appears today and again next week (relisted), Fredy may alert you again — deduplication is session-based, not permanent [1].

---

Who Should Use This / Who Shouldn’t

Use Fredy if:

• You’re actively apartment or house hunting in Germany and checking platforms manually is eating your day.
• You’re in a competitive city (Berlin, Munich, Hamburg, Frankfurt, Cologne) where being first to contact a landlord is the difference between getting a viewing and not.
• You have basic Docker familiarity or are willing to spend an afternoon learning it.
• You want real-time Telegram or Discord alerts without paying for platform premium tiers.
• You’re technically comfortable with the fact that scrapers break occasionally.

Skip it if:

• You’re searching outside Germany. The tool has no value for non-German platforms.
• You need guaranteed uptime and SLA-backed reliability. This isn’t that.
• You want AI-based filtering to score listing relevance — Fredy doesn’t do semantic matching, only platform-filter passthrough.
• You’re searching casually with low urgency and manual weekly checking is sufficient.
• You’re building a commercial product — re-read the scraping terms of service for each platform before doing anything business-facing.

---

Alternatives Worth Considering

Wohnungsbot — another German self-hosted apartment bot, older project, focused on Berlin specifically. Less actively maintained. Fredy covers more platforms and has a better UI [2].

Platform native premium tiers — ImmoScout24 Premium (~€40/month), Immowelt Plus, etc. These give you landlord-side priority contact features but don’t aggregate across competitors and have slower, batched notification delivery. Worth combining with Fredy rather than treating as alternatives.

ntfy + manual webhooks — if you have the technical chops to set up platform-specific webhook monitors yourself, this avoids the scraping fragility. Significant setup effort per platform; no deduplication.

Custom Python/Node.js scraper — the obvious DIY path. Fredy is essentially this, pre-built and packaged. Unless you have specific requirements Fredy doesn’t cover, there’s no reason to build your own.

Immoscraper — another open-source German apartment scraper project. Smaller, less polished, no Web UI. Fredy is the more mature choice for most users.

---

Bottom Line

Fredy is a narrow tool that does exactly one thing: it watches German real estate platforms so you don’t have to. The problem it solves — being first to see a new listing in a brutally competitive rental market — is real and the tool addresses it cleanly. Five platforms, six notification channels, deduplication, and a Web UI that doesn’t require reading documentation to use. The Apache-2.0 license and single-command Docker deployment remove most friction.

The honest caveats are equally real: scraping breaks, the ImmoScout24 integration operates in a legal gray area, and a single maintainer with 562 GitHub stars is not a team you’d bet an enterprise deployment on. For personal apartment hunting during an active search, none of that matters much. For anything beyond personal use, think carefully about the scraping legality under German law.

If you’re relocating to Germany, starting a job in Munich, or trying to get ahead of Berlin’s rental market, Fredy is the self-hosted answer to a subscription you’d otherwise pay €40/month for — and it covers more ground than that subscription anyway.

---

Sources

1. Fredy GitHub Repository and README — orangecoding. https://github.com/orangecoding/fredy (562 stars, Apache-2.0 license)
2. Fredy Official Website — fredy.orange-coding.net. https://fredy.orange-coding.net/ — includes feature overview, quick start, and live demo link