What Base44 Actually Is
Base44 is a vibe-coding platform that treats "app" as the deliverable, not a codebase. You describe what you want in a sentence; it returns a deployed, full-stack application - UI, database, authentication, hosting - that you refine through conversation. Founder Maor Shlomo bootstrapped it into one of the fastest-growing AI products of 2025, and Wix bought it mid-year for a reported ~$80M, which turned Base44 into both a standalone builder and the custom-app engine inside the Wix App Market.
The Wix integration is the differentiator that matters now. Open Base44 from a Wix dashboard and it detects what the site actually runs - Store, Bookings, Events, Restaurants, Blog, CRM - then proposes apps and agents built on that live data: a flash-sale campaign page that sells through Wix checkout, a profit-per-product dashboard, an agent that recovers abandoned checkouts or chases no-shows. Competitors generate generic apps; Base44 generates apps wired to a business that already exists.
Under the hood it runs a dual-credit economy: message credits buy your build-time conversation (25/month free up to 1,200/month on Elite), and integration credits buy your app's runtime - the AI calls, uploads, and database queries your end users generate (100/month free up to 50,000/month). Understanding that split is the difference between a $16 hobby and a $40 client-delivery business.
The Money Angle
- Client micro-apps for local businesses. The core freelance play: a clinic's appointment desk, a restaurant's lunch-special page, a salon's booking flow - one sentence to working software. Sell at $500-3,000 per build plus monthly care at $50-150, while your infrastructure cost sits on a $40 Builder plan. The approval-based workflow (build, demo, iterate) compresses what used to be a six-week agency job into days.
- Wix-ecosystem services. Millions of Wix sites run stores, bookings, and CRMs that the default dashboard doesn't serve well. Selling "custom apps and AI agents for your Wix business" - recovery agents, restock planners, staff schedulers on live data - is a new service category with a distribution channel no other vibe-coding tool has.
- MVP validation for micro-SaaS. Build the smallest working version on the free plan, put it in front of real users, and charge for it before writing production code. If it validates, either scale on higher tiers or rebuild on Replit/Lovable with an actual engineering budget - the $16 experiment cost is nothing.
- Internal tools for your own operation. The use case hiding in plain sight: profit dashboards, client-history views, restock forecasts built on your real data. Every hour saved is margin, and nobody bills you for it.
- Sell the build, own the support. Vibe-coded apps still need prompt maintenance as usage grows - and that maintenance is the retainer. The lock-in critics complain about is, from the service-provider side, the revenue guarantee.
What It Really Costs
| Plan | Annual (per mo) | Monthly | Message credits | Integration credits | Key additions |
|---|---|---|---|---|---|
| Free | $0 | $0 | 25 | 100 | Auth + database included |
| Starter | $16 | $20 | 100 | 2,000 | Unlimited apps, in-app code edits |
| Builder | $40 | $50 | 250 | 10,000 | Custom domains, GitHub, backend functions |
| Pro | $80 | $100 | 500 | 20,000 | Beta early access |
| Elite | $160 | $200 | 1,200 | 50,000 | Max limits, premium support |
The pricing trap is runtime, not build time. A client app with 200 daily users can burn a Starter-tier 2,000 integration credits on AI calls and uploads alone - the fix is pricing client retainers against usage tiers, the same way hosting companies always have.
The Honest Cons
- The free tier is a demo, not a workspace. Twenty-five messages end mid-project; plan on Starter the moment a build gets real.
- Lock-in is structural. Apps live on Base44's infrastructure; GitHub export (Builder up) gives you source, not self-hosting. Migrations are rebuilds.
- Usage-based costs scale quietly. Integration credits rise with your app's success while your plan price stays flat - monitor the meter as part of operating the app.
- The complexity ceiling is real. Prompt-built apps tangle at multi-role permissions and long feature lists; serious products eventually need engineering.
- Wix direction risk. The acquisition funds the roadmap - and could bend it toward Wix-site features over standalone builders.
vs The Field
| Base44 | Lovable | bolt.new | |
|---|---|---|---|
| Model | Sentence to hosted full-stack app | Prompt to React app, supabase-backed | Full-stack dev environment in browser |
| Free tier | 25 msgs + 100 integration credits | Daily message allowance | Limited daily tokens |
| Paid entry | $16/mo (annual) | ~$25/mo | ~$20/mo |
| Runtime cost model | Integration credits per usage | Included in tiers | Token consumption |
| Code access | Edits from Starter, GitHub from Builder | GitHub sync mid-tier | Full code by default |
| Distribution edge | Wix App Market + live Wix data | Design-forward community | Developer ecosystem |
Against the site's reviews: lovable wins for polished consumer-facing products with real design demands; bolt.new wins for developers who want code first and hosting second; Base44 wins for operators and freelancers whose product is "a working tool for a real business" - especially when that business already runs on Wix. Durable and 10Web (also reviewed here) build marketing sites; Base44 builds software.
Bottom Line
Base44 is the cheapest honest path from "this business needs a small app" to a URL that works - one sentence to full-stack, with the Wix acquisition adding a distribution channel its rivals cannot copy. The credit system is legible once you internalize the split: message credits are your building budget, integration credits are your operating budget, and client pricing should mirror both. The lock-in is real and the complexity ceiling exists, but for the micro-app and MVP tier where most side incomes actually live, $16/month for a hosted, login-ready product is the strongest cost-to-delivery ratio in the category. Start free, prove the idea in one sitting's worth of messages, then price the client work against the tiers.