Bubble is the most capable visual app builder available — real SaaS products, marketplaces and internal tools without writing code, now spanning web and native mobile. Its pricing has two layers and only one is on the pricing page. Web plans run $29 for Starter, $119 for Growth and $349 for Team, while the number that actually decides your cost is Workload Units: every database query, workflow and API call draws from a monthly allocation, from 175,000 on Starter to 500,000 on Team. The free tier is fine for learning and cannot publish to a custom domain.
Reviewed by EdGrows·Updated Aug 18, 2026Affiliate link — we may earn a commission
This analysis is based on documentation, public user reports, and vendor materials — not yet on our own hands-on testing. How we rate
Bubble publishes plan tiers. Your bill is those tiers plus something else entirely.
| Track | Range |
|---|---|
| Web | Free, Starter $29, Growth $119, Team $349 |
| Mobile | ~$42 to $449 |
| Web + Mobile | ~$59 to $549 |
| Enterprise | Custom |
Underneath sits Workload Units — a measure of the server processing your app consumes. Every database query, every workflow run, every API call draws from a monthly allocation: roughly 175,000 WU on Starter, 500,000 on Team, resetting each billing cycle.
Which means two apps on the same $29 plan can cost dramatically different amounts. A well-structured app with tight searches and backend workflows sits comfortably inside its allocation. A badly structured one — searches returning 500 records where 10 would do, workflows triggering workflows, scheduled jobs running hourly for no reason — burns through it and starts buying overage.
Experienced Bubble developers put the difference at 20–40% savings from query auditing and moving heavy logic into backend workflows. That's not a marginal optimisation; it's the gap between a viable cost base and an unpleasant one.
Worth stating plainly because it's the most common misunderstanding: the free tier cannot publish to a custom domain.
It's excellent for learning the editor, building a prototype and understanding whether visual programming suits how you think. It cannot host anything real. Launching requires Starter at $29 minimum.
That's a defensible design — Bubble isn't trying to be a free website host — and it does mean "try it free" carries an asterisk that competitors like Framer and Webflow don't have on their free tiers.
Bubble builds things the newer AI builders genuinely can't.
Multi-sided marketplaces with distinct user roles and permissions. SaaS products with subscription logic, usage limits and admin panels. Internal tools that connect to half a dozen APIs and enforce complex business rules. Real applications with real data models, running in production for years.
The plugin ecosystem matters more than it sounds — a decade of accumulated integrations means the payment provider, the mapping library, the calendar component you need probably already exists rather than needing to be built.
And the mobile track, added to cover native iOS and Android, closes a gap that used to force founders into hiring developers after all.
This is visual programming, not drag-and-drop.
You'll be designing database structures, writing conditional workflows, thinking about privacy rules and debugging why a repeating group returns the wrong records. There's no syntax to memorise and every other part of programming is still there.
People who arrive expecting Squarespace-with-logic leave within a week. People who accept it as engineering without code build things that would otherwise require a team — which is exactly why a substantial agency ecosystem exists around Bubble, and why "hire a Bubble developer" is a real job posting.
Budget several weeks to competence, not several evenings.
Bubble apps don't export to conventional code. The logic lives in Bubble's format, so leaving means reimplementing rather than migrating.
That's the standing objection and it's legitimate. If you expect this application to run for ten years and eventually need a development team, you're choosing a platform you'll one day rebuild away from.
The counter-argument that founders actually use: getting to market and finding out whether the product deserves a rebuild is worth more than architectural purity for a thing that might not survive its first year. Rebuilding a validated product is a good problem. Never shipping is not.
Compare that against Bolt or Lovable, which generate real framework code you own and can take to Cursor or any developer. They're faster to a first version and less mature for complex logic. Different bets, both reasonable.
Marketing websites. Wrong tool by a mile — Framer or Webflow.
Anyone wanting speed to a first version. Lovable and Bolt get there in an afternoon.
Teams needing code portability. No export path; plan accordingly.
Fixed, predictable budgets. Workload Units make costs behaviour-dependent rather than plan-dependent.
Against Lovable and Bolt: Lovable and Bolt generate application code from prompts — faster to something running, and the output is portable framework code rather than a proprietary format. Bubble has a decade of production maturity, a deep plugin ecosystem and better handling of genuinely complex logic. Speed and ownership versus depth and lock-in.
Against Softr: Softr builds portals and internal tools on top of existing data in Airtable or Sheets. Far simpler, far narrower. If your app is essentially an interface over a database you already maintain, Softr does it in a day where Bubble takes a fortnight.
Against Webflow: Webflow is a website platform with a CMS, not an application builder. They overlap only at the edges, and choosing between them means deciding whether you're building a site or a product.
Against Replit: Replit is a full cloud IDE with agents, aimed at people who will read code. Bubble is aimed at people who won't.
Against Supabase: Supabase is the backend that custom-built apps use. Complementary rather than competing — some Bubble apps connect to it for data that needs to live outside.
| Plan | Web | Mobile | Web + Mobile | Workload Units |
|---|---|---|---|---|
| Free | $0 | — | — | Limited, no custom domain |
| Starter | $29 | ~$42 | ~$59 | ~175,000 |
| Growth | $119 | — | — | Higher |
| Team | $349 | ~$449 | ~$549 | ~500,000 |
| Enterprise | Custom | Custom | Custom | Negotiated |
Checked August 2026 against current third-party verification of Bubble's published plans (annual billing, Web track). Workload Units are the primary cost variable and are consumed by database queries, workflows and API calls; extra units can be purchased, though consistently exceeding limits usually means upgrading. Plugin subscriptions, third-party API fees and custom domains sit on top. Confirm on bubble.io before committing.
Prototype on Starter and measure. WU consumption depends on how you build, not on what you build. Forecasting from a feature list guesses low every time.
Audit queries before scaling. Searches returning more records than needed are the most common cause of runaway consumption, and fixing them is reportedly worth 20–40%.
Move heavy logic to backend workflows. Same work, better economics.
Budget the ecosystem. Plugin subscriptions and third-party APIs are real line items nobody includes in the initial sum.
Bubble remains the most capable no-code platform for building actual software rather than websites. Multi-sided marketplaces, SaaS with subscription logic, internal tools wired into half a dozen APIs — these are things the newer prompt-to-app builders still struggle with, and Bubble has been running them in production for a decade. Adding a native mobile track closed the gap that used to send founders to developers anyway.
The pricing needs understanding before committing. Plans are $29 to $349 on the web track, and Workload Units decide the real bill — every query and workflow draws from a monthly allocation, so two apps on identical plans can cost very differently depending on how well they're built. The free tier can't publish to a custom domain, so launching starts at $29 regardless of what "free plan" implies.
Two structural costs sit outside the invoice. The learning curve is genuine — this is programming without syntax, and expecting drag-and-drop leads to abandonment. And there's no export path, so outgrowing Bubble means rebuilding rather than migrating.
For founders building a real product who will invest the weeks to learn it properly, recommend at Starter, then optimise before scaling. For marketing sites use Framer or Webflow; for speed to a first version with portable code, Lovable or Bolt.
Note: Bubble does not currently have an affiliate program with AIVario. We earn no commission, and this rating carries no commercial incentive.
Best for: Founders building marketplaces and SaaS products, complex application logic without a development team, internal tools with real business rules, teams that will learn the platform properly Not ideal for: Marketing websites, fastest path to a first version, projects needing code portability, buyers who want a predictable flat bill Bottom line: The deepest no-code platform available, priced by how efficiently you build rather than by what plan you pick — budget the Workload Units and the learning time, not the $29.
Web plans run from free to $349 a month — Starter at $29, Growth at $119, Team at $349. Mobile plans run roughly $42 to $449 and combined Web + Mobile plans $59 to $549. Enterprise is custom. Those figures are annual billing and they're only half the picture: Workload Unit consumption sits underneath every tier and is what separates a $29 bill from a $200 one on the same plan.
They measure server processing your app consumes — every database query, workflow run and API call draws from a monthly allocation that resets each cycle. Starter includes around 175,000 and Team around 500,000. Two apps on identical plans can cost wildly different amounts depending on how efficiently they're built, which is why experienced Bubble developers describe query auditing and moving heavy logic to backend workflows as worth 20–40% in savings.
No, and this catches people out. The free tier is genuinely useful for learning and prototyping but cannot host a live version on a custom domain or fully test API integrations. You need at least Starter at $29 to deploy anything real. Treat free as the classroom rather than the first rung — most teams move to Starter or Growth before onboarding a single user.
Build the prototype and measure, because prediction is unreliable. A search that returns 500 records costs far more than one returning 10, repeated workflows compound, and scheduled jobs run whether anyone is using the app or not. The practical approach is building on Starter, watching consumption on real usage patterns, then optimising before scaling. Teams that forecast from feature lists rather than from measured behaviour consistently guess low.
Yes — it added a mobile track, sold as separate Mobile plans or bundled Web + Mobile tiers. That's a meaningful expansion for founders who previously had to build a web app in Bubble and hire someone for iOS and Android. It also means three pricing tracks to choose between, and the combined plans carry a premium over web-only, so decide whether mobile is genuinely part of the product before paying for the track.
Steep, and the steepness is the point. Bubble is visual programming rather than a site builder — you're designing database structures, writing conditional workflows and thinking about privacy rules. People arriving expecting drag-and-drop simplicity bounce off it within a week. People who accept it as programming without syntax build things that would otherwise need a development team, which is why an entire agency ecosystem exists around it.
You rebuild, and that's the honest answer. Bubble apps don't export to conventional code — the logic lives in Bubble's own format, so migrating means reimplementing rather than porting. That lock-in is the standing objection to the platform and it's a real risk for anything you expect to run for a decade. For getting a product to market and finding out whether it deserves a rebuild, it's a trade many founders take deliberately.
Different bets. Lovable and Bolt generate application code from prompts, which is faster to a first version and produces code you own and can take elsewhere. Bubble gives you a mature visual environment, a large plugin ecosystem and a decade of production apps behind it, at the cost of lock-in. Prompt tools are better for speed and portability; Bubble is better when the app has complex logic and needs to be maintained by a non-developer.
The leading GTM data platform, repriced from the ground up in March. Two plans now, two credit types, and a lot of stale advice floating around.
The category leader in revenue intelligence, priced accordingly. The per-seat number is the smallest part of what you'll actually pay.
The best backlink data in the industry, sold with a credit meter and single-seat plans that make the real cost hard to read.