Is AI the Biggest Challenge Webflow Has Ever Faced?

I've been building with Webflow for about nine months now. It put my career on the map. I've shipped client projects, built my portfolio, and launched my own product — and Webflow was the tool I reached for every single time. So when I say what I'm about to say, know that it comes from someone who wants Webflow to win, not someone hoping it loses.

But I have to be honest: AI is the biggest challenge Webflow has ever faced. And it's not just AI — it's the entire ecosystem shifting underneath Webflow's feet.

Let me paint a picture. A developer wants to build a website today. Here's what they can do:

• Vercel for hosting and deployment — push to GitHub and it's live, free tier is generous, custom domains included
• Cursor or Windsurf for AI-assisted coding — write prompts, get production-ready components instantly
• Sanity for CMS — free plan, generous limits, fully customizable content models
• Supabase for database — PostgreSQL with auth, real-time subscriptions, free tier that gets you far
• v0 or Bolt for rapid prototyping — describe what you want and get a full UI built in seconds

This entire stack costs $0 per month to get started. And it scales. You're not locked into a visual canvas. You own your code. You can deploy anywhere.

Compare that to Webflow's Premium plan at $25/month per site for CMS access. Need analytics? That's another $9/month. Need localization? $9–29/month. Need optimization? $299/month. These add up fast.

Compare the modern AI stack — Vercel, Sanity, Supabase, Cursor — to Webflow's pricing. Zero dollars to get started versus $25/month just to unlock CMS. That math isn't doing Webflow any favors.

Here's a scenario I've seen multiple times: a client wants a simple marketing site with a blog. Maybe they'll publish 10 posts in the first year, maybe 20. They don't need 20,000 CMS items or 40 CMS Collections. They need a place to write and publish.

But to get CMS on Webflow, you're on the Premium plan. There's no "light CMS" tier. There's no "blogs only" plan. You buy the full CMS suite — the same plan a content-heavy publication needs — just to add three blog posts to your site.

Meanwhile, on Sanity, you can set up a blog schema in 10 minutes. The free tier gives you 1 million API requests per month. That covers most small-to-medium sites easily. And you're not locked into any single frontend — the content lives independently and can be consumed by any site, app, or platform.

If a client wants a marketing site with 3 blog posts, why should they pay for the same CMS plan as a publication running 10,000 articles? The math doesn't work.

Webflow's core value proposition has always been: build visually without writing code. Design, click publish, done. No development environment, no terminal, no npm install.

But AI is rapidly erasing that advantage. Cursor can build a fully responsive hero section from a single sentence. v0 generates complete landing pages from prompts. Bolt spins up entire apps — frontend, backend, database — in minutes.

A client who might have needed a Webflow developer two years ago can now type "build me a SaaS landing page with a pricing table, testimonials, and a dark mode toggle" into an AI tool and get something functional in under 60 seconds.

That doesn't mean it'll be good. AI-generated sites still lack design taste, accessibility considerations, and the polish that comes from a human who actually understands the brand. But the gap is closing — fast. And for budget-conscious clients who can't tell the difference between good design and functional design? AI is starting to look good enough.

To Webflow's credit, they're not sleeping through this. They've shipped:

• Webflow AI — built into the Designer, can generate sections, copy, and even entire sites from prompts
• Webflow Cloud — their own app hosting platform with databases (SQLite, key-value, object storage), essentially competing with Vercel
• Code Components — letting developers add custom React-like components into the visual canvas
• AEO (Answer Engine Optimization) — smart positioning against the AI search shift
• MCP server — exposing Webflow's capabilities to AI agents

These are solid moves. Webflow is essentially trying to become both a visual builder AND an app platform — a hybrid that lets you design visually while also hosting custom code, databases, and apps.

But here's the thing: Webflow is now competing on two fronts simultaneously. They're trying to match Vercel on the developer experience side while matching AI tools on the no-code side. That's an incredibly difficult position to be in, and the pricing model — designed for the old Webflow — doesn't fit the new reality.

I'm not here to predict Webflow's demise. I'm here to say what I think they need to do to stay relevant — and to keep developers like me building on their platform instead of jumping to the AI stack.

1. Fix the Pricing

The add-on model is hurting them. CMS shouldn't require the Premium plan. There should be a lightweight CMS tier — basic blog functionality, limited items, lower bandwidth — as a small add-on or even included at the Basic level. The current pricing pushes small clients toward alternatives before they even give Webflow a real chance.

2. Stop Charging for Essentials Separately

Webflow Analyze at $9/month on top of a site plan sends a clear message: "we're going to charge you for everything we can." Competitors bundle analytics. Vercel Analytics is free for the hobby plan and $10/month for teams. Webflow needs to stop treating basic analytics as a paid add-on.

3. Lean Harder Into the Visual Advantage

AI can generate code. AI cannot yet design with taste, understand brand identity, or make subjective creative decisions. Webflow's visual canvas — the ability to tweak spacing, typography, and interactions with direct visual feedback — is still unmatched. They should double down on this. Make the Designer so good that even developers who CAN code prefer it for certain types of work.

4. Own the Agency Workflow

Agencies and freelancers are Webflow's backbone. They should invest in multi-site management, client handoff workflows, white-label reporting, and better team collaboration at lower price points. The Team plan at $2,500/month is inaccessible to most small-to-medium agencies.

AI is a genuine threat to Webflow — but not because AI builds better websites. It's a threat because AI makes cheap alternatives viable, and Webflow's pricing makes those alternatives look even better in comparison.

I still build with Webflow. I still recommend it to clients who need visual control and don't have developers on staff. The platform is genuinely excellent at what it does. But I'm also building with Next.js, Supabase, and Sanity — and I'm watching AI tools get better every month.

Webflow doesn't need to beat AI. It needs to beat the value equation: make building on Webflow obviously worth the price compared to the alternative stack. Right now? That equation is getting uncomfortably tight.

Webflow doesn't need to beat AI at code generation. It needs to beat the value equation: make building on Webflow clearly worth the price compared to what you get for free elsewhere. That gap is narrowing — and pricing is the lever they need to pull.

If they fix the pricing, own the agency workflow, and keep pushing the visual canvas forward, Webflow stays relevant. If they treat every feature as a line item on the bill while the free alternatives keep improving? The developers who built their careers on Webflow might be the first ones to leave.