Comparing Vercel alternatives: can't tell half of them apart

This has been a productive but weird week. I'm ahead of schedule, more confident than ever in what my SaaS can deliver to the public, and about 80% sure I can build the native Mac app with the right features and infrastructure.

Read also: Mac app paused, recalibration, and content: update #12

But it's also been a week of asking myself questions I don't know the answer to.

Finding my weak spot

I've been going through my single points of failure lately. The parts of my stack where, if one thing breaks, everything breaks with it. Spacemail was one. Vercel is the other.

Read also: I dug into my other single points of failure after the Spacemail outage

Vercel is extremely reliable. I'm not questioning that. But reliable isn't the same as invincible, and outages can happen to anyone. I mentioned in an earlier post that I should probably go find some Vercel alternatives, just in case.

I'm not leaving Vercel. It's still my main driver. I just want to reduce the risk, and maybe move a small part of my future portfolio into a different deployment environment so I'm not fully dependent on one platform for everything I've built.

Asking ChatGPT, because obviously

I asked ChatGPT because I wouldn't know where to start on my own. I'd seen people on Reddit talk about different platforms, but I had no real sense of the trade-offs.

According to ChatGPT, Netlify is the closest thing to a Vercel alternative. Fine. But "closest alternative" doesn't fucking tell me what the real trade-offs are, and I still don't fully know.

Honestly, I barely know how Vercel works. My use of it is mostly the basics. Deployments, logs, build logs, that kind of thing. Don't ask me how it works under the hood or how to optimize any of it, because I can't tell you.

That's probably a mistake on my end.

No, it's definitely a mistake. I should be more diligent about understanding the infrastructure I'm building on, instead of just clicking deploy.

Understanding the core tech underneath my apps would probably make me better at writing prompts and instructions too, since right now I'm asking Claude Code to work with a black box I don't fully understand myself.

All the options

Netlify and Cloudflare are my first two bets. I've never used Netlify, but Cloudflare isn't a total stranger. I already use it for R2 storage on my docs editor and my dashboard. Not deeply familiar, but not unknown either.

Netlify looks like the easiest escape hatch. You connect the GitHub repo, it detects the build settings, and pushes trigger deployments.

Basically the same shape as what I'm already doing with Vercel. It's also started explicitly supporting AI-agent deployment with Claude Code and Codex, with Agent Runners that can make changes and generate deploy previews.

That's a critical detail to me, because Claude Code is already how I build everything.

Cloudflare is the more interesting long-term bet. It's not just "another Vercel".

It's Pages, Workers, R2, KV, D1, Queues, Cron, DNS, and CDN, all under one roof. But don't ask me how they perform or what each use is yet because I don't fucking know.

For a simple static site it's apparently straightforward: connect GitHub, point it at a build command and output directory, done.

That lines up with a concept I've had in mind, something similar to Stupid Hit, which was really just a JSON file and some basics.

Read also: $12, a typo, and the dumbest website I've built

Render, Railway, and Fly are complete dark horses to me. Netlify and Cloudflare are always the top recommendations on Reddit, so I'm mostly taking people's word for it there, but I've never seriously looked at the other three.

Maybe that's because they're less popular, or maybe because, from what I can tell, they're aimed more at experienced devs than at beginner vibe-coders like me.

Render stood out for one specific reason: it explicitly supports Claude Code, and it supports full-stack apps.

There's apparently a documented workflow where you install its deployment skill into your coding agent and just ask it to deploy.

I don't know what that means in practice yet, but it's the kind of thing that could fit how I already work.

Railway's the one that seems built for when an app grows past a simple frontend. Frontend, API, database, background worker, all in one project, which is more of a "future problem" than a today problem for me.

My understanding (so far)

I don't know. But if I had to guess:

Netlify is for simpler SaaS builds, which fits where I am right now. Render is about having more control over infrastructure once apps get heavier or more complicated. Cloudflare sits somewhere in the middle, with more infrastructure independence than the rest.

If I had to rank them today:

1. Netlify - easiest Vercel replacement

2. Render - particularly interesting because of the Claude Code support

3. Cloudflare - the best strategic hedge, and the deepest infrastructure ecosystem

What moves?

My code isn't going anywhere. It lives in GitHub, and that stays the source of truth no matter what I do with hosting. What I'd have to migrate is everything around the code:

• Build settings

• Environment variables

• Domains

• Storage and API integrations

• Deployment configuration

That's a much less scary list than "move my entire app somewhere else." Just like I was intimidated at first when migrating off my sites from WordPress to Github & Vercel.

Read also: Leaving WordPress After Twenty Years, in About a Day

What's my plan?

I'm not trying all three platforms at once. I'm an idiot, but not a total idiot. I can barely keep up with what's already going on. But I also don't want to get caught off guard if something breaks.

So the plan is simple:

• Build a small first project from my Idea Bench on Netlify

• If it's slow, annoying, or gives me too many headaches, move it back to Vercel, as a real test of what migrating looks like

• Once that's done, try Cloudflare and Render, one at a time, so I get properly familiar with each instead of half-learning three platforms at once

Read also: I added an Idea Bench to my blog

It's also a chance to learn how Vercel and Netlify really compare, instead of just repeating what Reddit or ChatGPT told me.

Why?

It's not some dramatic exit from Vercel. It's not "fuck you Vercel, I'm out." It's about reducing my risk of ruin. Making sure that if one platform has a bad day, my entire portfolio doesn't have a bad day with it. A hedge, not a breakup.

If deploying something new on Netlify goes well, I'll write about it. If I completely fucking cook myself trying, I'll write about that too. Either way, I'll write about it.