Articles · ComparisonsUpdated October 2026

The Bolt.new alternatives worth the move all answer one question: does your app have to leave the browser?

Most people searching for Bolt.new alternatives are building a web app, and the honest answer for them is to stay put. Bolt's help centre says it builds websites and apps, from marketing sites to dashboards and internal tools, and its pricing page puts hosting, custom domains and databases on the paid tier. Switching costs you a week and buys a different shade of the same thing. The reason to move is narrower: you need an app in a store, and you do not want the last step of the build to run on your laptop.

That last step is the whole story. Bolt builds mobile apps with Expo, and Bolt's own Expo guide tells you to download the code, install Node.js LTS and Git, install the EAS CLI, and run the builds from your terminal. Nothing about that is broken. It is just a different job from the one you signed up for in a chat box. Below are four tools people move to, every claim taken from the vendor's own page today. For the wider field we keep a running list of the vibe coding tools.

Jump to the shortlist

The short version

Bolt is a browser product, and the build step is where that shows.

Bolt's own glossary explains the machinery. It describes WebContainers as StackBlitz's browser based technology that runs Bolt's development environment, letting Bolt build, run and preview your app directly in the browser with no traditional server setup. That is why a Bolt project starts in ten seconds with nothing installed, and it is a genuine advantage for web work. It is also why the mobile path ends outside the tab.

So the useful way to sort Bolt competitors is by where the build runs, not by which agent sounds cleverer. Replit's docs describe a build that compiles on Expo's servers and lands in App Store Connect. Rork's docs describe a Swift app on a cloud iOS simulator and a build it submits to TestFlight for you. Newly runs cloud iOS and Android simulators and uploads to TestFlight and Google Play internal testing. Lovable and v0 keep you on the web, which is right if the web is what you wanted.

What running in the browser actually changes

For web work the browser model costs almost nothing. Bolt's supported technologies page says it focuses on JavaScript based web technologies, with Node.js for the backend, and states that languages such as PHP and Python are not compatible. Databases come from Bolt Cloud, or from Supabase through an integration. For a storefront, a dashboard or an internal tool, none of that is a real limit.

Mobile is where the model bends. Bolt's help centre says projects can be turned into mobile applications using Expo, and tells you to put the words mobile app in your very first prompt, because projects created for web do not easily switch over. Inside Bolt you scan a QR code into Expo Go to try the app on your phone. Then the guide hands you off: a computer with Node.js LTS and Git, an Expo account, an Apple Developer account, a Google Play Developer account, the EAS CLI, and the build and submit commands run by you.

Read that as a price rather than a flaw. If a terminal is already open, Bolt plus EAS is a sane pipeline and you keep the whole Expo toolchain. If the appeal was that you would never open a terminal, the handoff is where the product stops being the thing you bought. Whether an Expo project, a native project or a web page in a shell is the right output at all is a separate decision, and we set it out in web output versus native.

Bolt's own help centre, Expo for mobile apps

Four alternatives, and the reader each one is for

Lovable is the closest swap and the least likely to change anything. Its pricing page calls it an AI software engineer that lets anyone build for the web by chatting, and its free plan grants 5 build credits a day capped at 30 a month, plus 20 Cloud credits. Its docs list Git sync to GitHub, GitLab and Bitbucket on every plan, with code download on paid tiers. v0 from Vercel is the other near swap: a free tier with a 7 message daily limit, deployment to Vercel, GitHub sync and a visual Design Mode, with Plus at $30 per user a month. We ran one of those head to heads in lovable vs bolt.

Replit is the one Bolt users underrate. Its docs carry a full mobile track: choose Mobile app as the type, preview in an iOS Simulator or Android Emulator in the workspace, test on a phone through Expo Go, then use Launch in the Publishing tab. Replit's page says Launch creates the build, signs it, picks or creates the distribution certificate and provisioning profile, compiles on Expo's servers, lands it in App Store Connect, and can take up to 45 minutes. Its pricing page lists Core at $20 a month and Pro at $100, with no free tier shown.

Rork goes the opposite way from Bolt. Its docs say Rork writes native iPhone apps in Swift with SwiftUI and native Android apps in Kotlin with Compose, runs them on a cloud iOS simulator next to the chat, and submits to TestFlight and the App Store with no Xcode and no Mac. Web output is Vite and React. Its plans page lists Free as design mode only, Pro at $20 a month, and Max from $200 a month for iPad, Apple Watch, Vision Pro, games and widgets.

Newly sits next to Rork in intent and next to Bolt in language. You describe the app, it writes a React Native and TypeScript project you own, runs it on cloud iOS and Android simulators, uploads to TestFlight, publishes to Google Play internal testing and builds a standalone APK. It is $25 a month with no free plan, which is the clearest reason to pick something else: Bolt, Lovable, v0 and Rork all have a free tier you can open tonight.

Two questions

Which alternative is yours

What are you shipping?

Where should the build run?

Replit, Rork or Newly. All three document a build that runs on their machines. Bolt's own Expo guide sends you to a terminal with Node.js, Git and the EAS CLI at exactly this point.

What each one charges, from each vendor's own page

Bolt's pricing page, read on 1 October 2026, lists four tiers. Free is $0 with a 300,000 token daily limit, 1 million a month, Bolt branding on published sites, a 10MB upload limit and unlimited databases. Pro is $25 a month billed monthly, with no daily token limit, 10 million tokens a month to start, no branding, a 100MB upload limit, custom domains and rollover of unused tokens. Teams is $30 a month per member. Enterprise is quoted.

Comparing those figures across vendors is harder than it looks, because nobody sells the same unit. Bolt counts tokens. Lovable and Rork count credits, and Lovable's own page says a credit's value and burn rate differ by plan and by feature. Replit sells a subscription with a dollar allowance towards models. v0 publishes per million token rates by model. A headline price tells you the floor, not the bill. We found no published cost per finished app on any of these sites, and we are not going to invent one.

Price also hides what eats the most time on a first phone project, which is the accounts you open before a build exists. Bolt's Expo guide adds an Expo account on top of the Apple and Google developer accounts. Replit's Launch flow asks you to sign in with Apple. Rork's docs say new personal Google Play accounts need a closed test with 12 testers for 14 days. We wrote that paperwork down in vibe coding mobile apps.

Bolt's own pricing page, plans and token limits

What moving off Bolt actually involves

The good news is that Bolt does not hold your code. Bolt's GitHub integration keeps a full history outside Bolt, handles the Git steps for you, keeps each branch separate, and lets you build in Bolt, work in GitHub and come back. The Expo guide assumes you can download the project. Leaving is a repository transfer, not a rescue, and that is worth saying because plenty of alternatives to Bolt get sold on a lock in problem Bolt does not have.

The bad news is that a web project is not a mobile project. Bolt's help centre is blunt: projects created for web do not easily switch over to mobile. Carrying a finished Bolt web app into a mobile builder is usually a rebuild with your screens as the reference. So this decision is cheap before you start and expensive three weeks in. If the output is a website, stay. If it is an app in a store, pick the tool whose publishing step you are willing to perform, and read that vendor's own publishing page before you pay.

Bolt.new and four alternatives, on the things that decide it

ToolWhat the vendor says it buildsWhere the build runsPrice on the vendor's own pageStore route the vendor documents
Bolt.newWebsites and web apps; mobile apps via ExpoBrowser, on WebContainers; mobile builds on your own computerFree $0, Pro $25/mo, Teams $30/mo per memberDownload the code, install the EAS CLI, build and submit yourself
LovableWebsites and web appsBrowserCredit based; no dollar figure rendered to our fetchNone documented in its docs index; web hosting only
ReplitWeb apps, and mobile apps built with AgentBrowser; the iOS build compiles on Expo's serversCore $20/mo, Pro $100/mo; no free tier listedLaunch uploads to App Store Connect, then TestFlight
RorkSwiftUI iPhone, Kotlin and Compose Android, Vite and React webBrowser, on a cloud iOS simulatorFree is design only, Pro $20/mo, Max from $200/moRork builds and submits to TestFlight and the App Store
NewlyReact Native and TypeScript projects you ownBrowser, on cloud iOS and Android simulators$25/mo, no free planUploads to TestFlight and Google Play internal testing; you submit

Where we fit on this list

Newly is an AI app builder for phone apps. You describe the app in plain English and it writes a React Native and TypeScript project you own, runs it on cloud iOS and Android simulators while it builds, uploads the iOS build to App Store Connect and TestFlight, and publishes the Android build to Google Play internal testing while also producing a standalone APK. It does not submit for review. You press that yourself, from your own developer account.

New apps start local first, keeping their data on the phone with no accounts and no server. The agent adds Newly Backend only when the app needs accounts, shared data, syncing or payments, which brings sign in, an API service, a Postgres database per environment and file storage. Paywalls and subscriptions do come built in: ask the agent for one and it shows a Connect RevenueCat card. Code leaves as a ZIP from project Settings, through two way GitHub sync under Deploy, or with the CLI.

The reasons to pick one of the others are real. We are $25 a month with no free plan, so you cannot trial us the way you trialled Bolt. If you are shipping a website, Bolt, Lovable and v0 are built for that and we are not. If you want Swift on the phone, Rork writes Swift and we write React Native.

Questions people ask when they look past Bolt.new

It depends on what you are shipping. For another web app, Lovable and v0 are the near swaps and neither will feel like a step change. For a phone app you want built and uploaded without leaving the browser, Replit, Rork and Newly all document that path on their own sites. For a Swift project rather than a JavaScript one, Rork is the only tool here that writes Swift. If your project is a website and it works, the best alternative is usually no alternative.

Describe the app, then decide where the build should run

Write one sentence about what the app does. Then write a second about where the final build should happen: your terminal, or someone else's cloud. The second sentence picks the tool. If the answer is a phone app and a cloud build, start here.

Start building