best 16 ai app builders

prompt-to-app tools ranked on the two things that actually matter — can you own and export the code, and what the credits really cost — not on how good the first screen looks.

last reviewed 24 jul 2026 · 16 tools tested ·list curated by Onur Ozcanx in

the short version

best overallLovablefounders and developers building a serious mvp they may want to own later90/100 runner-upBolt.newdevelopers who want genuine ownership and a stack that isn't locked to one vendor89/100 best free optionv0developers who want top-tier ui and already live in the vercel/next ecosystem86/100

these are the 'vibe coding' tools: you describe an app in words — 'a crm with login and stripe checkout' — and get back a running, hostable application, not a code snippet. that is the line that separates them from ai coding assistants like cursor and windsurf, which live inside your editor and help a developer who already has a repo. app builders own the whole loop — generate, preview, provision a backend, deploy to a url — and assume you might not be a developer at all. (fully autonomous engineering agents like devin sit just outside this list too: they work tasks against an existing codebase rather than shipping a new app from a sentence.)

the single most important question in this category is ownership. some tools — bolt, softgen, v0, a0, same, emergent, lovable — let you export real source or push to github, so you can walk away and host it yourself. others keep your app on their platform forever: bubble exports no code at all, and the design-led builders are largely platform-bound. a beautiful app you can never leave with is a different purchase from one you own, and we rank the two differently on purpose.

the second trap is pricing. almost every vendor sells 'credits', 'tokens', 'messages', or 'effort', and the conversion to real work is deliberately fuzzy — a '$25/mo, 100 credits' plan tells you nothing until you know whether a credit buys one prompt or one-tenth of a feature. worse, token-based tools bill by codebase size, so cost climbs as your app grows, and replit's effort-based model has no pre-run estimate and no spend cap. we describe the unit each vendor actually uses and say plainly where the real cost is unknowable.

the third is demo-grade versus production-grade. many of these produce a gorgeous first screen and fall apart on the twentieth iteration — state, auth edge cases, database migrations. 'shipped a landing page' and 'runs a real business' are different claims, and the entries keep them apart.

1
Lovable
the market leader, with a real exit path off the platform
90/100
verdictthe best all-round balance of output quality, a real backend, and a genuine way out — two-way github sync plus the ability to migrate onto your own supabase. credit burn on heavy iteration is the tax.
best for
founders and developers building a serious mvp they may want to own later
price
$25/mo (Pro)
pricing note
free tier plus pro ~$25/mo and business ~$50/mo; the live page renders credit figures via javascript, so the dollar amounts are corroborated from multiple secondary sources — verify at checkout. credits roll over up to your plan cap and monthly credits expire after two months
free tier
yes
builds
full-stack web apps
output stack
react/tailwind/vite + supabase
code export
two-way github + zip
backend
supabase (provisioned)
pricing model
credits
lovable builds full-stack web apps: a react/tailwind/vite frontend on a supabase backend (auth, postgres, storage) it provisions for you as 'lovable cloud'. what pushes it to the top is the exit path — every save can push to your own github repo, there's a zip download, and in mid-2026 it added official buttons to export off lovable cloud onto a supabase project you control. you are not trapped, which is rare here.
the weaknesses are the category's: heavy iteration burns credits fast and the per-action credit value is opaque, and the stack is fairly fixed to react + supabase. commercially it's the clear leader — roughly $500m arr in 2026 after a $330m round at a $6.6b valuation — so platform risk is low. for most people building a real mvp, it's the default first pick.
pros
+two-way github sync plus a real migrate-to-your-own-supabase path
+full-stack: provisions auth, postgres and storage
+clear market and funding leader — low platform risk
cons
−credit burn on heavy iteration; per-action value opaque
−stack fairly fixed to react + supabase
−dollar pricing is javascript-gated on the live page
visit Lovable →Lovable alternatives →
2
Bolt.new
the developer's pick — real exportable code, flexible stack
89/100
verdictthe strongest ownership story with the most flexible output: it runs a real node toolchain in the browser, exports full code, syncs to github, and ships to netlify or vercel. token burn on big projects is the catch.
best for
developers who want genuine ownership and a stack that isn't locked to one vendor
price
$25/mo (Pro)
pricing note
free tier gives 1m tokens/mo (300k/day cap, no rollover); pro $25/mo starts at 10m tokens with unused tokens rolling over one extra month; teams $30/member/mo. tokens scale with codebase size
free tier
yes
builds
full-stack web apps
output stack
flexible (react/vite/next + supabase)
code export
download + two-way github
backend
supabase (your keys)
pricing model
tokens (scale with codebase)
bolt.new, from stackblitz, runs a real node toolchain in-browser via webcontainers, so the apps it builds use a flexible stack — commonly react/vite/next with supabase — rather than one fixed template. ownership is excellent: download the full codebase, two-way github sync, and one-click publish to netlify, vercel or render with your own keys. if leaving with real code matters most, it's the pick.
the recurring complaint is cost: because tokens scale with codebase size rather than message count, a growing project gets expensive fast, and the free tier's daily cap is tight. but for a developer who wants a running app and the freedom to take it anywhere, bolt is hard to beat.
pros
+full code export plus two-way github sync
+flexible, real node stack (not one fixed template)
+native netlify/vercel/render publishing
cons
−tokens scale with codebase size — big apps get pricey
−token burn is the top user complaint
−tight daily cap on the free tier
visit Bolt.new →Bolt.new alternatives →
3
v0
the best ui output, now a full-stack sandbox — on vercel's rails
86/100
verdictthe best-looking output in the category and, since early 2026, a real full-stack sandbox — but the iteration loop stays coupled to vercel and next, so you own the code less freely than with bolt.
best for
developers who want top-tier ui and already live in the vercel/next ecosystem
price
$30/user/mo (Plus)
pricing note
free tier gives $5 monthly credits and 7 messages/day; plus $30/user/mo, business $100/user/mo; model usage is token-priced. note some older listicles cite a $20 tier — the live page shows plus at $30
free tier
yes
builds
full-stack web apps
output stack
next.js + tailwind + shadcn/ui
code export
export + github (loop stays on vercel)
backend
supabase / neon / upstash
pricing model
credits + token model pricing
v0, from vercel, started as a ui generator and became a full-stack sandbox in february 2026, with api routes, server actions and database connections. its output quality — next.js + tailwind + shadcn/ui — is the best here, and it wires up to supabase, neon and upstash. you can export the code and import it to github.
the trade-off is ecosystem gravity: the build loop and full-stack sandbox are coupled to vercel and next, so owning it fully means taking the code local and out of that loop. if you're already on vercel and care most about front-end quality, it's excellent; if you want vendor-neutral ownership, bolt ranks above it for a reason.
pros
+best-in-class ui output quality
+now a full-stack sandbox, not just components
+tight supabase/neon/upstash integrations
cons
−iteration loop coupled to vercel and next.js
−less vendor-neutral ownership than bolt
−token-priced model usage on top of the seat price
visit v0 →v0 alternatives →
advertisement
4
Replit (Agent 3)
a real cloud ide you own — with the category's scariest bill
83/100
verdictthe most 'real dev platform' of the bunch — a genuine cloud ide with hosting where you truly own the code — held back by effort-based billing that can run away from you with no cap.
best for
developers and tinkerers who want a full environment, not just a generator
price
$20/mo (Core, billed annually)
pricing note
core $20/mo annual ($25 monthly) includes $25 monthly credits; pro $95/mo annual includes $100 credits. agent billing is effort-based — a simple edit ~$0.10, a complex feature '$5 or more' — with no pre-run estimate and no spend cap
free tier
yes
builds
full-stack web apps (+ more)
output stack
flexible (many languages)
code export
yes — full ide, github
backend
replit db / auth / storage
pricing model
effort-based credits (no cap)
replit's agent 3 builds full-stack apps inside a real cloud ide with databases, auth, object storage and one-click deployments, across many languages and frameworks. because it's a full environment, you genuinely own and can export the code or push to github — it's the least 'generator', most 'platform' option here.
the problem is the bill. effort-based agent pricing charges by how much work a request takes, bundled into checkpoints, with no estimate before it runs and no spend cap — the loudest complaint in the whole category, alongside repeated pricing changes. the platform is excellent; go in with eyes open about cost control.
pros
+a genuine full cloud ide, not just a prompt box
+you truly own and can export the code
+databases, auth, storage and hosting built in
cons
−effort-based billing has no estimate and no spend cap
−runaway-spend is the category's loudest complaint
−pricing model has changed repeatedly
visit Replit (Agent 3) →Replit (Agent 3) alternatives →
5
Emergent
a team of agents that ships web and mobile, and owns the code
81/100
verdicta fast-rising multi-agent builder that produces ownable react/next web apps and expo mobile apps, with real backend options — newer and less battle-tested than the leaders, but moving fast.
best for
founders who want a fuller multi-agent build, including a mobile app
price
$20/mo (Standard)
pricing note
free tier gives 10 credits; standard $20/mo (100 credits), pro $200/mo (750), team $300/mo (1,250 shared); credit-to-work conversion is opaque — confirm on the live page
free tier
yes
builds
full-stack web + mobile
output stack
react/next + node/fastapi; expo
code export
yes — ownable code
backend
node / fastapi + integrations
pricing model
credits
emergent runs a team of agents (architect, designer, developer, integration, pm) to build full-stack web apps in react/next with a node or fastapi backend, and mobile apps in expo/react native — one of the few here that does both, with ownable code and integrations like stripe on the paid tiers. the breadth is its draw.
it's newer and less proven than lovable or bolt, and like the rest its credit-to-work conversion is opaque. but the momentum is real — roughly $100m arr by early 2026 after a $70m round at a $300m valuation — so it's a credible, well-funded option for a fuller build.
pros
+builds web and mobile, with ownable code
+multi-agent workflow with real backends (node/fastapi)
+strong funding and momentum
cons
−newer and less battle-tested than the leaders
−credit-to-work conversion is opaque
−breadth means more surface area to go wrong
visit Emergent →Emergent alternatives →
6
Softgen
quietly the best value, with a full github repo per project
79/100
verdictthe underrated value pick: a yearly license plus wholesale pay-as-you-go credits, and a full unrestricted private github repo for every project. less polish and momentum than the leaders, but the ownership and price are excellent.
best for
budget-conscious developers who want real ownership without a subscription
price
~$33/yr + pay-as-you-go credits
pricing note
unusual model: ~$33/year platform license plus pay-as-you-go ai credits topped up to a wallet, no monthly fee; a 3-day trial follows a small ($3) credit purchase. effective per-build cost depends on usage
free tier
no
builds
full-stack web apps
output stack
next.js/ts/shadcn + supabase/firebase
code export
full private github repo
backend
supabase or firebase
pricing model
yearly license + pay-as-you-go
softgen builds full-stack web apps in next.js 15 + react + typescript + tailwind + shadcn/ui on a supabase or firebase backend, deploying to vercel. its two standout traits are price and ownership: instead of a monthly subscription it charges a ~$33/year license plus pay-as-you-go credits, and it gives you a full, unrestricted private github repo per project — the code is unambiguously yours.
the caveats are maturity, not capability: it's a smaller company with less community, polish and momentum than lovable or bolt, and pay-as-you-go can still add up on heavy use. for a developer who wants genuine ownership at the lowest realistic cost, it's genuinely underrated.
pros
+full unrestricted private github repo per project
+yearly license + wholesale pay-as-you-go, no subscription
+modern next.js/shadcn stack on supabase or firebase
cons
−smaller company, less polish and community
−no true free tier (small credit purchase to start)
−pay-as-you-go can add up on heavy use
visit Softgen →Softgen alternatives →
7
Base44
polished, batteries-included — and now owned by wix
76/100
verdicta polished all-in-one with a built-in database, auth and integrations — great for non-technical builders, but more platform-bound than bolt or softgen, and pricier since the wix acquisition.
best for
non-technical builders who want everything built in and don't need to leave the platform
price
$16/mo (Starter, billed annually)
pricing note
free tier gives 25 message credits + 100 integration credits/mo; starter $16/mo annual ($20 monthly), builder $40, pro $80, elite $160. dual credits: message credits to build, integration credits for runtime ops; users report ~15-30% higher effective cost post-wix
free tier
yes
builds
full-stack web apps
output stack
own all-in-one platform
code export
platform-bound (verify)
backend
built-in db + auth
pricing model
dual credits (message + integration)
base44 builds full-stack web apps on its own all-in-one platform, with a built-in database, auth and integrations rather than a bring-your-own stack — which makes it fast and friendly for non-technical builders. it was acquired by wix in june 2025 for around $80m plus earn-outs, and hit roughly $100m arr within nine months, so it's well-resourced.
the trade-offs are ownership and cost. it's more platform-bound than the export-first tools — confirm the current export options before assuming you can leave with your code — and long-time users report effective costs 15-30% higher since the wix deal, partly due to the dual message/integration credit system. batteries-included, at a price.
pros
+polished, batteries-included: built-in db, auth, integrations
+friendly for non-technical builders
+well-funded (wix-owned), strong momentum
cons
−more platform-bound — confirm export before committing
−effective cost rose ~15-30% post-wix acquisition
−dual-credit system complicates cost estimates
visit Base44 →Base44 alternatives →
8
Tempo
planning-first: diagrams before code
73/100
verdicta yc-backed, planning-first builder that maps user flows and architecture before writing react — a disciplined approach that reduces slop, though it's smaller than the leaders and the assisted tier is enterprise-priced.
best for
teams that want a disciplined design-then-build workflow rather than instant slop
price
$30/mo (Pro)
pricing note
free tier gives 30 credits; pro $30/mo (150 credits); a human-assisted 'agent+' tier is $4,500/mo. confirm on the live page
free tier
yes
builds
full-stack react apps
output stack
react (planning-first)
code export
react code (verify path)
backend
full-stack react (verify provider)
pricing model
credits
tempo builds full-stack react apps but leads with planning: its agents produce user-flow diagrams and an architecture before generating code, which tends to reduce the incoherent output that plagues one-shot builders. it's yc-backed (s23) with a $5m seed, and generates react you can take.
it's smaller and less proven than the top tier, its human-assisted 'agent+' tier is priced for enterprises, and you should confirm the export path and backend provider on the live docs. for a team that values a design-then-build discipline, the approach is a genuine differentiator.
pros
+planning-first (diagrams + architecture) reduces slop
+react output you can take with you
+yc-backed with a disciplined workflow
cons
−smaller and less proven than the leaders
−human-assisted tier is enterprise-priced
−verify export path and backend on live docs
visit Tempo →Tempo alternatives →
9
Same.new
cheap, honest token ladder — great for drafts and ui clones
71/100
verdictthe cleanest, cheapest token ladder in the category, with downloadable projects and a knack for cloning an existing ui from a url or screenshot — but a thinner backend story, so treat it as a drafting tool.
best for
solo builders who want fast, cheap web drafts or to clone an existing ui
price
$10/mo (Basic)
pricing note
free tier gives 500k tokens/mo; basic $10/mo (2m), pro $25/mo (5m), max $50/mo (10m), ultra $100/mo (20m + pay-as-you-go). tokens do not roll over — they expire each cycle
free tier
yes
builds
web app drafts + ui clones
output stack
react-family web
code export
download on paid tiers
backend
light
pricing model
tokens (no rollover)
same.new turns prompts into web app drafts and is unusually good at cloning an existing ui from a url or screenshot. its pricing is the most transparent here — a simple token ladder from $10 to $100/mo — and paid tiers let you download projects and remix, so you can take the code.
the honest limit is depth: its backend story is lighter than lovable or bolt, and tokens don't roll over, so it sits between a no-code prototyper and a code-first builder. for cheap, fast drafts and ui clones, it's excellent; for a complex full-stack product, look higher.
pros
+cleanest, cheapest token ladder in the category
+download and remix projects on paid tiers
+excellent at cloning a ui from a url/screenshot
cons
−thinner backend story than the leaders
−tokens don't roll over — they expire each cycle
−more draft/prototype tool than production platform
visit Same.new →Same.new alternatives →
10
Rork
the best-known prompt-to-mobile-app builder
69/100
verdictthe best-known way to build a cross-platform mobile app from a prompt, now with a native-swift option — but mobile store submission still has real friction, and you should confirm the export path.
best for
non-technical founders shipping a mobile app without touching xcode
price
$25/mo (Junior)
pricing note
free tier gives 35 credits/mo (5/day cap); junior $25/mo, middle $50, senior $100, scale $200-$1,800. a native-swift 'rork max' tier is $200/mo. confirm on the live page
free tier
yes
builds
mobile apps (ios + android)
output stack
react native; native swift (max)
code export
verify on live docs
backend
external
pricing model
credits
rork is the best-known prompt-to-mobile builder: it generates cross-platform ios/android apps in react native (and, via 'rork max', native swift compiled on a cloud mac fleet), with a path to publish on the app store and google play. for a non-technical founder who wants a mobile app without learning xcode, it's the obvious starting point.
the caveats are mobile-specific: app-store submission and review still carry real friction that no builder removes, the free tier's daily cap is tight, and you should confirm the current export path on the live docs before assuming you can take the project. within mobile, it's the category leader by name recognition.
pros
+best-known prompt-to-mobile builder
+cross-platform react native plus a native-swift option
+path to app store and google play
cons
−app-store submission friction remains
−tight daily cap on the free tier
−verify export path on live docs
visit Rork →Rork alternatives →
11
a0.dev
cheap prompt-to-mobile that hands you a real expo project
67/100
verdictthe developer-friendly mobile pick: it generates a standard, exportable expo project you can open in vs code — genuinely not locked in — for $20/mo. it's early-stage and small, but the ownership story is clean.
best for
developers who want a cheap, portable react native starting point
price
$20/mo (Pro)
pricing note
free tier plus pro $20/mo; confirm current tiers on the live page
free tier
yes
builds
mobile apps (react native)
output stack
expo / react native
code export
download expo project
backend
external
pricing model
subscription
a0.dev builds mobile apps in react native and its real strength is ownership: it generates a standard expo project as real .tsx files with hooks and navigation that you can download and open in vs code — you are not tied to a0's environment. at $20/mo with live previews and a path to the app stores, it's a cheap, portable starting point.
it's early-stage — a small ($500k) seed in 2025 — so weigh longevity, but for a developer who wants a real, exportable expo project rather than a locked mobile platform, it's the cleanest option at the price.
pros
+generates a standard, exportable expo project
+open it in vs code — genuinely not locked in
+cheap at $20/mo
cons
−early-stage and small
−app-focused — external backends
−less polish than the market leaders
visit a0.dev →a0.dev alternatives →
12
Riff (formerly Databutton)
python-backed app building, now aimed at the enterprise
64/100
verdicta capable react + python builder with a real backend, now repositioned for enterprise governance after rebranding from databutton — solid technically, but pricing is a moving target we couldn't verify.
best for
teams wanting a python backend and governed, enterprise-oriented deployment
price
not confirmed post-rebrand
pricing note
databutton became riff in october 2025 and pivoted to enterprise/governance; the live riff pricing page could not be confirmed at review time, so we don't publish a figure — the old databutton tiers are stale. verify before relying on any price
free tier
no
builds
full-stack web apps
output stack
react + python (fastapi)
code export
verify current path
backend
neon postgres + auth
pricing model
credits (unconfirmed)
riff — the former databutton, rebranded in october 2025 after a $16m series a — builds full-stack web apps with a react frontend and a python (fastapi-style) backend, plus neon postgres and auth, and has repositioned toward enterprise and it-governed deployment. the python backend depth sets it apart from the react/supabase crowd.
the honest problem is the pivot: it's a moving target, and its current pricing page could not be confirmed at review time, so we don't quote a number — the old databutton tiers are stale. verify pricing and the current export path directly before committing. technically sound, but in transition.
pros
+react + python (fastapi) backend depth
+neon postgres and auth built in
+enterprise/governance positioning
cons
−current pricing unverified post-rebrand
−product in transition after the pivot
−no longer aimed at solo builders
visit Riff (formerly Databutton) →Riff (formerly Databutton) alternatives →
13
Trickle
simple, cheap and predictable — for lighter apps and sites
61/100
verdicta simple, affordable builder with predictable pricing for websites and lighter apps — good for straightforward projects, but not for heavy full-stack work, and it's a small company.
best for
non-technical users who want simple, predictable pricing for lightweight apps
price
$20/mo (Pro)
pricing note
free plan plus pro ~$20/mo (custom domains, larger databases) and premium ~$50/mo; some sources list the tiers slightly differently, so confirm on the live page
free tier
yes
builds
websites + lighter apps
output stack
trickle-hosted
code export
leans platform-hosted (verify)
backend
hosted db
pricing model
subscription
trickle builds websites and lighter apps with a refreshingly simple, predictable pricing model — a free plan plus roughly $20 and $50/mo tiers that add custom domains and larger databases. for non-technical users who want something straightforward without credit anxiety, that clarity is the appeal.
it leans platform-hosted, so confirm the export path if ownership matters, and it's a small company (roughly $1m raised) building lighter-weight apps rather than complex full-stack products. within that scope, it's a clean, cheap choice.
pros
+simple, predictable pricing
+cheap, with custom domains and databases on paid tiers
+friendly for non-technical users
cons
−lighter-weight — not for heavy full-stack apps
−leans platform-hosted; verify export
−small company (~$1m raised)
visit Trickle →Trickle alternatives →
14
Bubble
a mature no-code engine with ai — but zero code export
58/100
verdicta mature, genuinely capable no-code platform that added an ai generator — but it exports no source code at all, so your app is locked to bubble permanently. that single fact caps where it can rank here.
best for
non-technical builders who accept lock-in for a battle-tested visual engine
price
$29/mo (Starter)
pricing note
starter $29/mo (10,000 workload units), growth $149/mo (30,000 wus), team $349/mo (50,000 wus); 14-day trial requires a card. workload-unit pricing scales with app usage
free tier
no
builds
full-stack web apps (no-code)
output stack
proprietary visual engine
code export
none — permanent lock-in
backend
built-in db + auth
pricing model
workload units
bubble is a decade-old no-code platform with a real, battle-tested visual engine — built-in database, auth and workflows — and it added a 'start with ai' generator that produces a working prototype in minutes plus an ai agent that edits your app. for genuine application complexity without code, the engine is proven.
the decisive weakness in this category is ownership: bubble exports no source code, so your logic lives on bubble forever unless you rebuild elsewhere, and it isn't self-hostable. add workload-unit pricing that climbs as your app grows, and the ai generator reads as a capable bolt-on to a platform whose lock-in is the real story. that's why it ranks where it does despite being technically strong.
pros
+mature, battle-tested visual engine
+handles real application complexity without code
+built-in database, auth and workflows
cons
−no source-code export — permanent platform lock-in
−not self-hostable
−workload-unit costs climb as the app grows
visit Bubble →Bubble alternatives →
15
Framer AI
gorgeous ai-built websites — but sites, not apps
54/100
verdicta design-led site builder with genuinely great ai-generated output — but it makes marketing websites, not full-stack apps, so it's here as the boundary of the category rather than a true app builder.
best for
designers and marketers who want beautiful sites fast
price
$10/mo (Basic)
pricing note
basic $10/mo, pro $30/mo, scale from $100/mo (billed annually), editor seats +$20/mo; free tier includes 500 ai credits/mo. ai = wireframer site generation + workshop components
free tier
yes
builds
websites (design-led)
output stack
framer-hosted
code export
framer-hosted
backend
none (site tool)
pricing model
subscription
framer's ai (wireframer for whole-site generation, workshop for components) produces beautiful, design-led websites quickly, and for designers and marketers it's a strong, affordable tool. we include it because people reach for it in the same search, and to draw the line clearly.
that line is: framer builds sites, not full-stack apps — there's no real backend or business logic, and it's framer-hosted rather than a code-export app builder. judged as a prompt-to-app tool it ranks low, not because it's bad, but because it's a different thing done well.
pros
+gorgeous, design-led ai site generation
+affordable, with a real free tier
+great for marketers and designers
cons
−builds websites, not full-stack apps
−no real backend or business logic
−framer-hosted, not a code-export app builder
visit Framer AI →Framer AI alternatives →
16
Webflow AI
ai-assisted sites and cms — still a site tool, not an app builder
52/100
verdicta deep, design-led site and cms platform with an ai assistant and claude mcp integration — powerful for content-heavy sites, but like framer it's a site tool, not a prompt-to-app builder.
best for
teams managing and optimizing content-heavy marketing sites
price
$29/mo (Standard site plan)
pricing note
site plans standard $29/mo, plus $74/mo, advanced $212/mo; standalone ai-assistant pricing isn't separately published. ai = ai assistant, claude mcp integration, aeo tooling
free tier
yes
builds
websites + cms
output stack
webflow platform
code export
static markup (higher tiers)
backend
none (site/cms tool)
pricing model
subscription
webflow pairs its mature site and cms platform with an ai assistant, a claude mcp integration and answer-engine-optimisation tooling, giving teams deep control over content-heavy marketing sites. for site management and content ops it goes further than framer.
but it's the same boundary case: webflow builds websites and cms-driven content, not full-stack apps with real backends. it exports static markup on higher tiers, yet judged as an app builder it sits at the edge of the category — included for completeness and to keep the definition honest.
pros
+deep site and cms management
+ai assistant plus claude mcp integration
+static-markup export on higher tiers
cons
−builds sites/cms, not full-stack apps
−no real application backend
−standalone ai pricing not clearly published
visit Webflow AI →Webflow AI alternatives →

how this ranking was made

prices come from each vendor's own live pricing page where it could be read. several render their prices through javascript credit calculators that can't be fetched — lovable is the notable one — and where that happened we say so and use figures corroborated across multiple independent sources rather than a number we couldn't see. verify any price at the logged-in checkout before relying on it.

we do not publish an effective cost-per-app or cost-per-feature, because for most tools here it genuinely isn't knowable — the credit-to-work conversion is unpublished, and replit's effort billing has no pre-run estimate at all. we describe the pricing unit honestly and stop there.

code export and ownership is the highest-stakes fact on each entry and the most likely to change, so we treat it as a first-class spec and state exactly what you can take with you. where a vendor's export path wasn't fully confirmable, the entry says to check it before committing.

acquisitions, rebrands and shutdowns are dated and sourced: base44 was bought by wix in june 2025, databutton became riff in october 2025, and google's firebase studio is being sunset in march 2027 with new workspaces already disabled — so it isn't ranked.

funding and arr figures come from press and tracker sites and are directional, not audited; they inform the 'is this company going to be here next year' read, not the ranking's core.

our general methodology and disclosures →

was this useful?

Lovable

Bolt.new

v0

Replit (Agent 3)

Emergent

Softgen

Base44

Tempo

Same.new

Rork

a0.dev

Riff (formerly Databutton)

Trickle

Bubble

Framer AI

Webflow AI

how this ranking was made