Anything specific to watch out for?

On what url were those? Or Cloudflare's TOS in general ?

I just spent a week writing a harness around the existing tunnels to make this by hand.

No, they claim their edge is better for residential connections to your backend. We tried multiple SF to us-east-1 and it was always substantially worse

This requires no port forwarding, so someone brute force scanning your IP won't find this and your open port. they also might have a harder time figuring out what service has been exposed since the port number itself will be unknown.

They're scanning all IPs all the time. If your home router has a public IP and a firewall log, check it

This is a crowded space with lots of solutions (oss and commercial)

https://github.com/anderspitman/awesome-tunneling

More like their version of Tailscale funnel I think

But you do seem to get to host a https version of your app in case you need features locked behind secure context.

That can't be right. If they're hosting on a different DNS they have an absolute need to MITM ssl/tls traffic. Can't work otherwise.

So cloudflare sees your plaintext. Btw: tailscale does not (but ssl errors and warnings are unavoidable)

It's not new. Maybe this page is new, but being able to set up a quick tunnel on trycloudflare.com without an account has been a thing for years.

There's pros and cons on the longevity side. 24/7 use causes more wear to mechanical parts (fans for instance) but can be better for the electronics, as heat cycles are a big part of the problem there.

Very few people own desktops anymore. Generally people only buy desktops if they need to run a GPU, such as for gaming or running a local model.

Yeah my desktop is on 24/7. Laptop is on when in use. Phone is on all the time too, of course, but I don't use it for browsing or work nearly as much as my desktop.

There are people who turn off their computer?

You do if you're hosting a website on it :-)

Maybe it's a nice way to quickly test a new service or change over the internet without going through the git process, same way one would use ngrok - I could see that being helpful when prototyping / pocs, etc - before deploying changes to your app via the proper channel, especially if you are already using cloudflare for your domains/apps. As other people said, it probably would not be practical or scalable for most people to run an app 24/7 from their laptop using their home or office internet connection.

This is pretty great and I think this will be quite important in the age where everyone has their self-hosted services.

Gotcha, that totally makes sense.

The "centralization" which is cloudflare basically running its own walled garden version of the interent (how often do you see a cloudflare page checking if you're human?) is exactly why a lot of people do NOT like cloudflare. And if you've known about CF and its leadership since their inception you'd be even more wary of sticking your stuff over there.

I've migrated many companies off of cloudflare, usually because they end up pissing off companies when a contract renewal comes up and they slam them with massively increased bills and almost useless support if you aren't very high paying enterprise. I don't know how many CF support tickets I've just given up on over the last 15 years, usually related to their admin page, workers or some weird thing their system does that wasn't documented and I just stop getting responses and definitely don't get fixes.

If you ever worked in webhosting the Cloudflare wordpress/etc extensions are everywhere and back when I did work in hosting tons of support tickets were made because of CF. Could be way better now, I don't go near that industry these days.

The casual CF user sticking it in front of a blog they rarely look at and the business forced CF user has a very different experience. I cringe and seriously consider if I'm interviewing for an infra role and they use cloudflare. Usually it's startups that grew into larger businesses.

Don’t the free tunnels have explicit limits on bandwidth and streaming?

Because it only matters whether people actually switch away from them.

I didn't take it as a capacity concern, but a "how much data do they get to look at" one.

Just use TLS / mTLS over the tunnel, no?

I think they're talking about market capture risks

Sure any centralization of infra is an obvious risk for a myriad of reasons eg DoS, manipulation, honeypots, etc.

But again my point applies - the chances they get enough people using it that it becomes a meaningfully worse security target than lots of other existing things seems … super low.

It’s a big world, people make many choices I can’t understand (nix? Haskell? Php? <flame wars to /dev/null>). Even if this product nailed it - the number of people who can use it is minuscule - yes even as we add Claude-enabled PMs to the software dev ranks.

Alt view with the old saying - “put all your eggs in one basket … and watch that basket!”

True. Same reason Hurricane Electric peers promiscuously. I'm surprised more networks don't, to be honest - wouldn't say DTAG prefer that you peer with DTAG than peer with HE upstream of DTAG?

Interesting how 4/4 other replies didn't get the centralisation concern despite it being a fairly often discussed topic

Yeah, it reminds me of Google. Fool me twice ...

I'm not trusting any of these corporates any more

There are other reasons not to trust them as well, especially if you've seen how their sales team operates https://news.ycombinator.com/item?id=40481808

Your opening piqued my interest, but:

“The substrate itself consists of a few systems…”

I doubt that this is how wordy your communication is.

“It consists of a few systems” would be adequate. And if we had prior context about what else exists that surrounds “the substrate” the “substrate itself” distinction would be meaningful, but it’s not, because you are referring to one object, which is the system you built, and I doubt any enzymes act on it, so it’s likely not a substrate.

Spain blacklisted Cloudflare and everyone hates it, but it proves it can happen.

Wow, exfiltrating data has never been easier!

How does “open sourcing the tech” “give nefarious actors on the internet more opportunities”? Sounds like a paltry excuse for not open sourcing your tech.

Probably why they used the trycloudflare.com domain as they expect it to be blocked.

I don't know if there's an overall limit, but their regular tunnels have a limit of 100 MB per request which breaks stuff like Immich.

These are the limitations mentioned on the docs [1]. Quick Tunnels are subject to a hard limit on the number of concurrent requests that can be proxied at any point in time. Currently, this limit is 200 in-flight requests. If a Quick Tunnel hits this limit, the HTTP response will return a 429 status code. Quick Tunnels do not support Server-Sent Events (SSE).

[1]: https://developers.cloudflare.com/cloudflare-one/networks/co...

My AI discovered this days ago when I wanted to deploy a new vibe coded website (it was to keep score while playing whist and rentz)

Thought it was very cool

hobbiest or dev just testing your services

Is different than sharing your work with a client.

I personally wouldnt make it SOP to utilize a complely free service like this to share my work. I'm not saying it's not convient, it definitely is.

But you shouldn't subject a clients product to terms they probably aren't aware of.

It's the same reason that exists for PoW captcha, it's just way more annoying. Also accounts offer traceability versus a new identity every time you want to do something unless you want to go through hundreds of accounts a day.

yep, and all those fired and remaining are busy ai-native proofing their careers and tokenmaxxing everything including copy

I swear if I see one more "pill-badge callout -> header -> subtitle" with gradient background and hover-cards landing page I'm going to crash out

AI took all the effort, so I guess nobody sees value in these little pages anymore. Not even those being paid. It's actually scary.

I have a proxy set up with the Oracle Always Free VMs they provide. It works well enough.

For free idk, I personally use a VPS (with unlimited traffic) and Tailscale. Someone has to pay for the public IP and proxy.

i just use wireguard or tailscale, provision a subdomain with a custom ssl and resolve to a nat ip.

Use pangolin (you can self host), been using it for a while and it’s great, under the hood it’s a vpn+reverse proxy which you can do yourself too. In pangolin you can have public or private resources, where private ones you need to authenticate through pangolin first (either pass or others like pin number for your parents so easy to remember). When you link your domain for public ones, I suggest you make a sub domain for it, so your apps will be a sub to your subdomain, that way you keep control of your main domain while having automatic assignment for your apps rather than manually, and if you didn’t issue a certificate, that sub.subdomain is basically invisible on the internet unless you host a service to expose it.

I know. I'm not "upset" over it, just that it makes the service less useful for me.

You can put nginx with basic auth in front of jellyfin

We are in the age of where hype beats reasoning.

I see, good point. Some might be confused about my comment then.

which proves the point OP is making.

I'm viewing the site on lightmode, clicked one of the #links which scrolled down the page, when I scrolled back up all the sections lost their content and were unusable.

I've never seen someone fail a static page like this. And from a huge company like Cloudflare too..

I’m raising my eyebrows…

Cloudflare is going full vibecoding for at least one year. They are high on AI psychosis.

[ad]

frp is nice. I have the tiniest Amazon Lightsail instance running a tunnel to a Minecraft server in my basement. It’s cheaper than getting a fixed IP from my ISP, at sufficiently low traffic.

The website looks broken in dark mode on Firefox.

You should be more than hesitant - don’t do it. They literally market them for quick demos. And not long lived production.

I use for the exact same thing. Running "cloudflared tunnel" sets up a temp solution. No log in needed, pretty convenient.

From their docs [0]:

Free tunnels are meant to be used for testing and development, not for deploying a production website.

[0: https://developers.cloudflare.com/cloudflare-one/networks/co...

As far as I can tell, the difference is that they made a new website? The cloudflared instant tunnels have been around for years, including json output as far as I know.

This gives a random trycloudflare.com subdomain each run, the other requires a domain and tunnel key on your account.

It looks like tunnels can do some non https stuff these days? But it's a bit unclear

Cool but the obvious flaw with this is that CF leaks DNS-records. So bots will find these urls instantly. Not sure why they have not fixed that or if it is even possible to fix.

Love this, very cool.

I used to use a service called ngrok for this, but it's nice that Cloudflare is offering one now.

how long until someone's agent sets up a tunnel for the world to see one's most sensitive, private and embarrassing information or insecure work-in-progress app? granted, for the brew install some massive permissions are needed, so hopefully for those running their agent's in a sandbox, you should be safe.. should be..

So... Cloudflare built Ngrok?

That's handled by all the prior releases of cloudflare tunnels

ngrok does this free, but with an account

I've been running a couple of websites* on a home machine behind a Cloudflare tunnel for 2-3 years now. The only thing I had to pay for was the DNS registration. There are limits on the traffic you can serve on their free plan, but I'm nowhere near reaching those. I don't know what an SDK would add, you just set up a web server using whatever method suits you, then configure the cloudflared to connect to that. I did find that the online dashboard could not support all the options I wanted, so I switched to using the config file method - it would have been easier if I used that method from the start.

One caveat - there is an option to inject Javascript into your pages for traffic measurement which is ON BY DEFAULT, you have to go to "Web Analytics" and turn this off if you don't want it!

* https://cmacleod.me.uk & https://newsgrouper.org

Only reason I have it (cloud flare tunnels) setup for personal use is because I had to do it for work years ago and yeah, Enterprise(tm) is a good description, haven't touched it in years so when it breaks I'll likely use something else.

Getting it to do what I wanted with a traefik front router with cloudflared talking to arbitrary subdomain that is spun up and broadcast from the other project side was very painful in a "this could be more friendly" way.

It has been truly bulletproof though since then so that's almost enough to make me go reread all the docs.

If only headscale supported tailnet lock and multiple control servers. They consider both features "enterprise" even though the former is arguably a bare minimum and the latter nice even for home users in the event their lab gets taken offline.

So you have a phone to phone VPN and each runs services or just a regular rapid/synology at home with containers?

I’m not GP, but run something that’s pretty close to GPs description. It’s been a fun journey to set up.

The core is chat+audio/video call server running locally on my m3 Mac Studio. Centrifugo has handled the chat concerns very well. LiveKit was a really nice foundation for video and audio calls. There are a few different options for local STT if you want that.

I used RN via expo for the client and have my friends and family on TestFlight as beta testers. Utility over polish.

With messages and call transcripts on my own box, I can prompt Claude code or Codex to operate on any message or transcript content. And follow up in chat with a message. I do most of that from Claude or ChatGPT mobile apps via remote control to my sessions running in the box.

From there, if I give enough of a specc, anything that happens in chat or call transcripts can become an additional custom workspace in the main app. I think GP calls these mini apps. But they’re essentially rich clients under the main app umbrella. There’s other details, but yeah it’s a strange new world. Check my profile and reach out directly if you want.

Not OP- I think there’s been some cool solutions popping up around hosting personal apps. On the front page this week was Capsule (https://withcapsule.app/). My personal solution is Exhibit (https://github.com/momja/Exhibit) which also has the goal of being a self-hostable artifact store.

I’ve been building something similar with [Homestead](https://myhomestead.dev). It’s my OSS solution for building mini apps. Each mini app is a Resct frontend and the backend (storage, auth, MCP, API, etc) just comes for free.

I have my own solution here that also works for sharing stuff with clients with configurable whitelists at artifacts.iofold.com (also open-source for self hosting on cf workers for almost free, github linked on the homepage)

I run something similar, and in my system an app artifact is typically just a rich react component in a React Native client. They usually start as prototypes in chat, similar to what the labs provide in their native apps. If I or someone else in my user group wants more depth than a chat widget, then the artifacts can grow into a “workspace”. Oddly enough my richest workspaces have most been for entertainment. DnD, MtG deck building, HN comment thread parsing. And system management. Intrusion detection, feature life cycle/LLM handoff management, call telemetry. I do have a workspace for public county permit lookup which has been handy too. Reach out and I can share more.

That's why I keep using tsnet directly, where you can have as many virtual machines, doing sharing or funneling, as you want to.

The CLI doesn't let you do as much as the go library does.

You can set up paths that route to different local ports. As long as your apps don't need to live at the root path you can have as many as you like. And you can still have something on the root path as long as it doesn't need the subpath.

Here's `tailscale funnel status` on my machine:

    $ tailscale funnel status
    
    # Funnel on:
    #     - https://my-machine.tailXXXX.ts.net
    
    https://my-machine.tailXXXX.ts.net (Funnel on)
    |-- /            proxy http://localhost:3000
    |-- /foo         proxy http://localhost:3001
    |-- /bar         proxy http://localhost:3002
    |-- /baz         proxy http://localhost:4004

I have multiple funnels running at once. I just put a tailscale container into my compose stack and that way each compose stack gets its own tailscale instance and thus its own funnel. Happy to share the setup if at all interested.

Reverse proxy to additional endpoints.

Also a very good alternative. I went on a spree and tried out all of the self hostable alternatives and this was definitely a runner up to netbird. The self hosted tailscale is also good, just not as polished as these two projects are.

Which features do you specifically benefit from?

I think Tailscale would've fit my needs perfectly fine, but the management makes access control and grouping dead simple in netbird. It also generates a nice little map of your peers, their groups, and connected nodes etc. Quick visual on your topology.

What ssh features? You can have a hundred devices connected for free and they can all just ssh to each other directly

Both support multi-user setups with SSO or built-in auth.

Beyond that, compared to a typical hub-and-spoke WireGuard setup, the main advantage is peer-to-peer connectivity. Clients connect directly to each other when possible, which lowers latency by bypassing a central relay.

AFAIK, they also have different origins:

Pangolin started as an internet-facing reverse proxy (Traefik) combined with a WireGuard server for backend nodes. It has gradually added VPN-like features, including client device access and an internal HTTPS proxy similar to Tailscale Serve.

NetBird is a self-hostable Tailscale alternative that started as a mesh VPN focused on P2P traffic. It recently added its own reverse proxy features (Traefik-based, coincidentally), also similar to Tailscale Serve.

Pangolin is centered on endpoint and ingress management, while NetBird focuses on mesh networking, though their feature sets are increasingly converging.

I pay Tailscale nothing, why switch?

I think an important feature for homelabs is NAT traversal

No, it’s not. It handles the keys, provisioning, DNS, NAT traversal, and a bunch of other stuff. WireGuard is a great technology - Tailscale is like a usability layer on top of it.

Wireguard is great for a point-to-point or multipoint VPN set up by a competent network administrator between machines with static addresses. But that's the only thing it does. It does not handle authentication or mobility very well.

git : github :: wireguard : tailscale

I use tailscale on my steam deck to play stardew valley with my son on my laptop, it truly is amazing

Strange, battery use is negligible on Android, particularly when the tunnel is not in active use.

I have the same experience and I once found a GitHub issue about it. I don’t think it sounded like they planned to fix it anytime soon. I just added the Tailscale toggle button to my phone’s quick control panel and turn it on whenever I want to check all my syncthing nodes or something. I keep meaning to see if I can script up something with Shortcuts to make it turn back off automatically.

WG gonna punch through NAT for me?

My endpoint is a pretty stable (though technically dynamic) IPv4 on one end, but the other might be a cell phone with CGNAT, some random WiFi, blah blah etc. TS does that. If you don't want to use it, cool. Don't. I'm willing to make the tradeoffs to use TS for now. That could change in the future.

I wasn't, no. I'm not the only one whos hit it either, not sure what the cause is :(

Using Tailscale in exactly the same way. It has been such a blessing!

Fyi you can point a domain/subdomain (eg lan.mysite.com) at that 100.x IP and it will only work for those on your tailnet.

I use both (also Headscale)

I just use /remote-control. You can't do everything from remote but it's was easier just opening the Claude app than using ssh. And I don't need to be on my tailnet either.

If by local MCP you mean stdio, then a tailnet won't help you unless you proxy the local MCP using something like mcp-proxy: https://github.com/sparfenyuk/mcp-proxy . That local mcp is only exposed to apps on the machine. The flow is Phone client --> agent harness or whatev --> Tailnet --> Mac Mini --> mcp-proxy http listener --> local mcp. But you need to wrap the local mcp runner in the proxy (check the docs).

EDIT you don't mean running on your actual phone lol, you mean in the web browser at Claude.ai. Anyway, if you use mcp-proxy for a stdio mcp, or if it's an http mcp alone, then you would need a public endpoint for Claude.ai to connect to. Tailscale will only help you if you use Funnel, but this cloudflare thing is exactly what you need (w/o Tailscale).

I built something for this exact use-case using Cloudflare Tunnels:

https://github.com/dalberto/mcp-ferry

I also use CF managed auth to make auth easier.

it’s missing the bigger integration offerings Tailscale has/does still.

Until I had tailscale serve generating valid certs, I had a good reason to use Cloudflare tunnels.

But in general I don't want to put everything on the internet side of things.

Mostly, I don't want something open, but more like a "share with" for people who are in the same office (virtually over tailnet, not physically on the same LAN).

This still works great for a demo instead of a product pitch, to send an link out to see something.

I'd still use a real host over a laptop for those.

That's really cool! I really want something like that. Care to share what your stack is? My wife would love having that super power.

Don't think we had any issues with Warp but Cloudflare APIs are a bit of a mess. Seems like they've cobbled together and then renamed/rearranged things multiple times now and it's a bit hard to keep up.

Terraform has worked decently well especially since there's a few random settings here and there that aren't exposed anywhere in the UI (facepalm)

Think shared Claude Artifacts that don’t live @ Anthropic.

Shameless plug, I've built a self-hosted capability here with things like live collaboration for humans and agents. There is a native cloudflare deplyoment and integration with Cloudflare Artifacts. PR for tunnel would be appreciated.

https://github.com/plannotator/artifact-server

I have really enjoyed the experience of OpenZiti for this purpose. Being able to host dark services and impose granular access controls and routing for different clients is a game changer.

For the few services I host that require SSL (eg; WebUSB), I serve the dark service via a standard domain (example.com) so I can still get a LetsEncrypt cert, but public access to that domain resolves to a static page; "Plz connect to OpenZiti & try again". This allows me to have SSL on required dark services without requiring to install a private CA for each client.

I use NextDNS instead of public DNS to do the same thing.

I would say that I've been doing most of that already with a reverse proxy+SSO/IDP for a very long time. Not the same, but close-enough and open to the "traditional internet" in a manner that does not require my users to install and configure tailscale, which is a massive plus. I do appreciate tailscale for the option to keep my effective host IP address hidden/local, but then I've been having a front on the internet (with a public IP/domain/...) for decades and I don't see a problem continuing.

How do I get this more persistent?

I never found cloudflare appealing for the small developer.

Now that is one vibe-coded website

Thanks for sharing.

- Co-founder of Pinggy.

You can also do this on any server you have:

ssh -R 0.0.0.0:80:localhost:9051 myserver

Make sure server sshd config has: GatewayPorts yes

The number of vibe coded apps accidentally hosted on dev laptops is about to explode.

This is not anything new they have been around since the tunnels project launched. What is more concerning is the AI slop webpage and just the "AI" centered relaunch of an existing product.

So is this basically a ngrok from CF?

As I understand it you can make a tunnel, but it's under a domain you control and it's a few clicks in the dashboard to set up, and then a something to run in CLI. That's what my homelab uses. This is just a single CLI command and the whole thing is set up.

Works on someone's computer.

It's "free". That's all you need to understand.

As far as I know this has been a provided feature of cloudflare for several years. I don't understand why it was posted today.

The page is a fine exemplar of marketing deception. Banner says:

Free, secure tunnel for everything you are building.

Preview and ship ideas globally in seconds with Quick Tunnels. Deploy your local application to the Internet with a single command.

Clicking through Explore Cloudflare Tunnel leads us to:

Looking to expose public applications? This documentation covers Cloudflare Tunnel use cases for private networking and Zero Trust, like VPN replacement and private network access. For publishing public web applications, APIs, and services to the Internet through Cloudflare refer to ...

Oh I use this for testing and demos between my phone and computer. Ngrok is fine, probably better off using Tailscale, but it's very easy to just run a quick tunnel with "cloudflared tunnel --url localhost:XXXX".

Off-topic: That web page looks exactly like my vibe-coded utilities.

Cloudflare's only goal here is to put themselves between everyone and charge a toll. They are running a protection racket where they themselves are causing much of the damage they offer protection from. This is another example.

So this is basically ngrok?

So it is ngrok 10 years later?

Group of three cards, each with a left-aligned icon, heading and subheading stacked -- that's the m-dash of web design.

I'll take that over corporate memphis of the previous decade.

(idk if it's really declining or I simply haven't noticed it in a while)

It routes from closest PoP to visitor -> closest PoP to origin. Your way doesn't convey that as clearly IMHO.

Claude has a habit of always talking in an overly active voice, sort of personifying subjects that feels off. I don't think many people would choose to say "Traffic rides" here.

Thanks! The build system is nix, but the result is more of an appliance than a distro. (linuxManualConfig from tinyconfig + fragment, static musl Rust PID 1, mke2fs -d) Updates are rebuild-redeploy with the provider API; it can't update itself. The filesystem is a fixed-size image (even omitting resize2fs, so no growing onto the VPS disk), and it remounts read-only after boot. The boot log is 270 kernel lines then 4 userspace: 1 nftables loaded, 2 printing the gate's wg0 public key, 1 remounting ro.

Self hosting might win on control, but it’s harder to manage so I went managed (CF Tunnel). Edge terminates TLS, my Caddy only speaks HTTP, nothing to cert renew (unless you cerbot, but that’s not a guarantee if Let’s encrypt is down).

No, by definition it's not P2P because it's all going through Cloudflare.

I'm sorry to be offtopic, but this website looks like an hallucination generated by chatgpt 4.

This type of quality downgrade is scary, and it's everywhere now.

cloudflare rocks

This has been the behaviour when you launch but don’t connect a specific tunnel for at least a couple of years

Quick tunnels have existed since at least 2022 (when I was using them)

Nothing is new, quick tunnels were launched in 2021

https://blog.cloudflare.com/quick-tunnels-anytime-anywhere/

I want this with TOTP or passkeys. Just giving me a hostname doesn't cut it these days.

Tailscale is one of the most profoundly useful technologies I’ve ever used as small business person. It’s also one I would have never thought to invent despite feeling its lack daily.

tredre3 explained it above

precisely (thanks!)

cloudflared is a bit unique in that it makes multiple outbound connections to CF DCs to do the tunneling and fails over if a DC has an issue (or, more likely, when a DC is near-capacity and they need to divert lower-tier traffic away from it[0]). SSH would reintroduce a single point of failure to this.

Also, the _main_ use case of `cloudflared` tunneling is using it as a long-term way to host production websites on your own hostname. the ability to create ad-hoc tunnels is more of a gimmick / advertising opportunity.

0: https://github.com/judge2020/cloudflare-connectivity-test/wi...

*60 minutes

I still remember when ngrok came out. The experience was more or less the same.

It’s interesting that 10 (more?) years later the product has not evolved and, apparently, hasn’t found a way to finance itself without removing the pure free tunneling option.

How many ways can you convey doing the same thing? All the banks out there have almost identical slogans because they do very similar things.

No, marketing is about communicating valuable offers.

Misleading communication is what I am describing, that is a subset of all communications.

The main use case: cloud harnesses to access your local network; I highly recommend to run the harness in your local network instead, then you own the session, no tunnel needed.

The other use case for webhooks is ok, but is exactly what ngrok already does since forever with a pretty high free quota.

Seems like the quality of software engineering is now going to zero and just vibe coding unstable alternatives since 2023.

This is cool. I like it! Also Cloudflare design has stepped up so much over the past 2 years.

Does this make working with agents in cloud more interesting? Maybe now I can see my changes right away without waiting for the build and deploy? By running the dev server and tunneling from the sandbox giving me a direct url access somehow?

Can someone chime in here?

Ah yes. We can't let self hosted app traffic flow across the Internet in a way that escapes the vast Internet surveillance system run by the Americans called Cloudflare.

Hmm, Tailscale is too convenient and that traffic is going dark from Cloudflare's all seeing eye.

It's the AI witch-hunt these days. Salam trails coming soon to you're nearest artist! "That's AI!!!"

Sigh... HN littered with it these days.

Would you ever ship a product where all the ad copy was written by an AI and read by no one?

The marketing from Cloudflare around this being for agents feels like the intended use case is for agents to exfiltrate data off your laptop without pesky hurdles like asking the user to set up an account first.

It's not a security issue, it's a liability issue!

I'd expect an anonymous inbound tunnel to be better than an anonymous outbound proxy and we have plenty of the latter.

json output is nice, my agents always had to parse the text output before

This is my docker compose for sharing files with randoms online https://pastebin.com/raw/cYV7ksw9

Don't pass -d, your tunnel URL prints in the console and you can download dirs as tarball.

Needs to support tcp and udp

I have been considering using my other laptop as a place to run my agents in production since 1. my laptop is more powerful than the vps I can rent for cheap.

then just use Cloudflare tunnels to connect to the laptop.

For small-scale personal or family applications serving just a few users, does it make sense to choose Quick Tunnels over Tailscale or Pangolin?

It's broken on desktop too...

This is super neat! I rely on normal Cloudflare Tunnels for my self-hosted website/web apps, works shockingly well all things considered.

It was pretty easy to setup… but I cheated, and use Dokploy which handled most of it AFAICT

I used Cloudflared tunnels for a while in production, they worked very well but the config was a bit of a pain sometimes. It's nice to see a simpler version even if it's just for testing.

I'm not sure how safe this is...

vibe coded app with may be no security and now available from the internet for anyone to RCE into my laptop?

I’ve been using cloudflare tunnels quite a bit in places where I would have traditionally gone for a reverse port forward and they’re working really well. The quick tunnel concept seems really great but the non-quick tunnels are also very useful and can be entirely automated.

I did not know about them and I have a good use for them today. So I found it useful though I agree it needs the date disclaimer

Didn't these used to be called Argo Tunnels a long time ago? I think Argo Tunnels got repositioned for connecting to origin servers without needing an incoming connection to the origin but I thought they originally did something like this/ngrok

Yeah, I just checked the page >can connect their server to the Internet with Argo Tunnel for free

Why is it more relevant today?

I saw the post and was literally thinking "I have been using these tunnels extensively for a couple years now"

The punch and profound-sounding phrasing is also Claude; but pre-LLM it might make sense for a marketing page.

"Your laptop stays private. The URL goes everywhere."

A URL ... goes, does it?

I've seen it pointed out (& have noticed) that anthorpomorphising like this is an LLM smell.

In this case, although "URL goes" can be valid, it's just awkward here.

One tell is: if you can code that hero there's no excuse for the silly bugs.

For example, the hero is fairly impressive in mobile but it's broken in desktop. And then you scroll down to the next section and there's overflowing text.

IOW, the skill level is inconsistent.

Yet so many developers do not know about it.

They are useful for taking local llm models and delivering them with HTTPS to harnesses that expect it or to another machine. Other than that I'm not sure what I'd personally use them for.

i find it hard to read any text when the site has many animations that can't be stopped

I so hate codex speak.

I've used these tunnels all the time when I wasn't at my computer. Insanely helpful when doing agentic work on the go

perhaps this is the most easy to use networking product that ever shipped, but given ngrok did this many years before, maybe they just think for the agentic era? though one must take care for the authentification

I also think it seems cloudflare enter into a stage keep adding products and make the portal looks like a maze to get more advanced features configured

Do you still remember cloudflare mesh and use and stick to it?

I have been developing for my own use something like that but self hosted. There are other solutions, but this one simple, reliable and secure in a way I see it.

Originally didn't want to share but here it is https://github.com/reddec/tunnel-me

UI: fully LLM assisted (not vibe coded, but guided with a lot of iterations). Backend: hand written, but before release polished via LLM. Docs: me - input, LLM - output.

The things I am proud:

- it very reliable

- its single binary with reasonable defaults and low memory usage

- SSO out of the box (cause I am using pocketid in my homelab)

- very simple backend

Cloudflare tunnels are really useful when dealing with webhooks. Way better than Ngrok and you can set it up on a custom subdomain you own. eg: site.yourdomain.com

Lol Codex UI!!!

I have used this extensively previously when I had to quickly share a UI prototype with my Product manager. There used to be all sorts of cors error when I also had a backend. Then I realised its just better to make a branch and deploy it in amplify and add the cors origin in the backend. But its good for quick UI sharing

I use this a lot, but find it a bit hard to manage if you want a persistent tunnel locally to use for dev (e.g. webhooks etc), so I added a wrapper around it that you can easily use: https://github.com/cliftonc/cfld

The RAM pricing crisis is doing wonders to the software industry. Amazing.

What could possibly go wrong? ;-)

I have been using CF tunnels for quite some time, very useful if you want to ingest mail on a local server without a public IP and hosting anything locally.

They are providing this for free up to a point, I wonder if all developers are aware of this amazing offering?

Lovely to see Cloudflare Tunnel frontpage on HN. Working on Tunnel was the highlight of my career. It was a fun product with a lot of bugs to fix, features to add, architecture to design. We were all working in-person, solving incidents together, getting lunch together every day. Taught me so many important skills that have helped me in every other project I've ever worked on.

Ah, an ngrok competitor. It's what we elderly (~40 myself), yelling-at-clouds types call "reverse SSH tunnels", but this is an anachronism from the era of knowing how to do things rather than paying for managed SaaS products to do them. Don't mind me, just yelling at clouds...