How much does it cost to move my Bubble app to custom code, and how long will it take?
Short answer: Moving a Bubble app to custom code is a rebuild, not a file export, because Bubble does not support code exports. Bubble is a no-code tool where you build apps by dragging pieces together. For a small app with a few screens, plan on roughly 6 to 12 weeks of developer time. A larger app with many workflows and years of data takes longer. The cost is hours times a rate, so use the line items below, and only move if your monthly Bubble bill or a missing feature makes the payback clear.
You are thinking about leaving Bubble and want a real range for the cost, the time, and the point at which it pays for itself.
I'm Ranjit, and I run Blur Studio: 15+ years building software, 120 apps taken live, and clients who have gone on to raise $10M+. I build AI systems and processes that help businesses do a lot more with a lot less time. I also work differently from most of what else is out there, more on that further down this page. Say hello.
The short answer
You cannot press a button and get your Bubble app as code. Bubble's support centre says it does not support code exports. You can export the app as a JSON file, a structured text file, but it is made for importing into another Bubble app. Your data can be exported as CSV, a plain table file, from the editor.
So moving means a developer rebuilds your app in code, using your Bubble app as the specification, and then moves your data across. The price is the hours that takes. I give the hour ranges below with a stated rate, so you can replace my assumptions with the quotes you get.
There is no single market average for this, so the figures here are planning assumptions, labelled as such.
What you pay for in a migration
You are not paying to copy. You are paying for five separate jobs.
- Reading your app. A developer has to understand every page and workflow. Bubble workflows are the step-by-step rules behind each button. A long list of them is the biggest driver of time.
- Designing the database. Bubble hides the database from you. In code it has to be designed on purpose.
- Rebuilding screens and logic. The work most people picture.
- Moving the data and the users. Records, files and logins.
- Testing against the old app. Running both side by side until the answers match.
Cost by app size
These are my planning ranges, in developer weeks and hours, with the cost at an example rate of $100 an hour. The hours and the rate are assumptions for illustration, so replace $100 with the rate you are quoted. Figures are as of November 2026 and in US dollars.
| App size | What it looks like | Planning hours | Cost at $100 an hour | Timeline |
|---|---|---|---|---|
| Small | 5 to 10 pages, one user type, a few dozen workflows, no payments | 200 to 400 | $20,000 to $40,000 | 6 to 12 weeks |
| Medium | 10 to 30 pages, two or three user types, payments, emails, an admin area | 400 to 900 | $40,000 to $90,000 | 3 to 6 months |
| Large | 30 or more pages, many integrations, heavy data, years of history | 900 and up | $90,000 and up | 6 months or more |
At $60 an hour every dollar figure drops 40%. At $150 it rises 50%. If a quote falls far outside the range for your size, ask which line items are different from the list below.
The line items behind those ranges
| Line item | What drives it |
|---|---|
| Discovery and written spec | Number of pages and workflows. Expect 20 to 60 hours. |
| Database design and build | Number of data types and how tangled they are. |
| Login and accounts | Bubble handles sign-up for you. In code it is built or bought in. |
| Screens | Count of unique screens. Repeated layouts are cheaper. |
| Business rules | Each workflow with conditions, loops or scheduled jobs. |
| Integrations | Payments, email, maps, anything through a Bubble plugin. A plugin is an add-on someone else wrote. Each needs a replacement. |
| Data migration | Rows, files, and how clean the data is. |
| Testing and parallel run | The riskier the app, the longer both versions run. |
| Hosting and live setup | Domain, backups, error alerts. |
Data and login risks that add cost
- Your data comes out as CSV, not as a database. Relationships between records, such as which order belongs to which customer, have to be rebuilt carefully and checked.
- Files and images are stored separately and need their own move.
- Users and passwords. Plan for every user to set a new password unless your developer confirms another route. Ask Bubble's support before you promise your customers anything.
- Who can see what. Bubble controls who sees which data inside the platform. In code the same limits must be written again. If your new backend is Supabase, a database platform, the Row Level Security rules do this job, and they need testing. See the security checklist for what to test.
- Scheduled jobs and emails are easy to forget because they run in the background.
What Bubble costs you today
To judge payback you need your real monthly number. Bubble's pricing page, read in November 2026, lists these web plans, billed annually, per month: Free at $0 with 50,000 workload units, Starter at $32 with 175,000, Growth at $119 with 250,000, and Team at $349 with 500,000. Enterprise is a quote. A workload unit, or WU, is Bubble's measure of the server work your app does. Extra file storage is $3 per 100 GB a month. If you use more WU than your plan includes, you can buy a workload tier or pay overages as you go. Bubble says it emails you at 75% and 100% of your workload. Plans may be priced differently if billed monthly or with mobile, so check the page.
My longer guide, Bubble agency vs custom code, covers how to cut that bill without leaving. Do that first if your workflows are slow or wasteful.
A payback example with made-up numbers
Say a small business owner runs a customer portal on Bubble. All of these numbers are invented to show the arithmetic.
| Item | Amount |
|---|---|
| Bubble plan plus overages, per month | $700 |
| Hosting, database and tools after the move, per month | $100 |
| Monthly saving | $600 |
| Migration cost (a small app at $100 an hour, 180 hours) | $18,000 |
| Payback | $18,000 divided by $600 equals 30 months |
Thirty months is a long time. Now change one number. If the Bubble bill is a steady $250 a month and the new running cost is $100, the saving is $150, and payback is 120 months. That is ten years, so you would not move for money reasons.
My rule: if payback is under about two years, the move can pay for itself on cost alone. If it is longer, you need a reason that is not on the bill, such as a feature Bubble cannot do, a customer who demands you own your code, or a plan to sell the business.
A hybrid option: move in phases
You do not have to move everything at once. Three phased routes cost less up front and lower the risk.
- Move the heavy part only. If one feature, such as a report, a calculation or an AI step, eats most of your workload, rebuild only that as a small service in code and have Bubble call it. Check Bubble's documentation for the connection options.
- Move the data first. Put the data in a database you control, keep Bubble as the screens for now, and replace screens one area at a time.
- Rebuild one user group. Move the admin tools or one customer type, then the rest.
The trade-off is that you pay for both systems for a while. For the wider comparison of Bubble with the newer AI builders, see Bubble vs AI app builders and alternatives to Bubble.
What about FlutterFlow and other builders
FlutterFlow is a tool for building mobile and web apps visually. Unlike Bubble, its documentation describes pushing your project to a GitHub repository, a place that stores code, so you get real code to hand to a developer. That makes a move from FlutterFlow a different, usually smaller, job: a developer takes over the code instead of rebuilding from a description. If you are on an AI builder, the guide on exporting code from an AI app builder shows what you get.
Who should do this, and who should not
Move if your bill is rising faster than your revenue after tuning, if you need something Bubble cannot do, or if owning the code matters to you. Stay if the app works, users are happy and the bill is steady. Then the money is better spent on a good Bubble developer. A fractional CTO, a part-time technical leader, can help you decide without committing to a build; fractional CTO vs development agency explains the difference.
Whoever you hire for the move, ask for a written spec of your current app first. If they cannot describe what your app does before quoting, they cannot price it.
How I'd approach this
I would start with the bill, not the code. Pull the last six months of Bubble invoices and see whether the number is steady or climbing. If it is steady and the app works, I would not move.
If it is climbing, I would first pay a Bubble expert for a day to cut workload. That is cheap and tells me whether the platform is the problem or the workflows are.
If I still wanted to leave, I would pay for a written spec and a fixed quote on phase one only, usually the data and the heaviest feature. I would never sign one price for the whole rebuild before a developer had read the app. Then I would run both versions side by side until the answers match.
Blur Studio fixes and finishes start around $1,500, and a Bubble move is scoped on your actual pages and workflows.
Questions people ask
Can I export my Bubble app as code?
No. Bubble's support centre says it does not support code exports or on-premise self-hosting. You can export the app as JSON for use in another Bubble app, and export your data as CSV. Moving to code means a rebuild. My guide to Bubble agency vs custom code covers the alternatives.
How long does a Bubble migration take?
In my planning ranges, a small app takes about 6 to 12 weeks, a medium one 3 to 6 months, and a large one 6 months or more. These are estimates built from hours, not a promise. Ask each developer for dates by phase.
Will my users lose their accounts?
Plan as if they will have to set new passwords, and ask Bubble's support and your developer what can be moved before you tell customers. Move the account records first, test login, then send a clear email.
Is it cheaper to rebuild in an AI app builder than in custom code?
Sometimes for a simple app. For anything with payments, private data or many rules, I would have a developer check the result either way. The comparison page Bubble vs AI app builders goes through it.
Can I keep Bubble and still cut costs?
Often yes. Bubble charges by workload units, so slow or repeated workflows cost money. A Bubble expert can tune them. Compare that cost with the migration payback shown above before you decide.
What should I hand a developer before they quote?
Your Bubble editor access (view only is enough to start), a list of pages and user types, the plugins you use, your last six months of Bubble invoices, and a CSV export of your main data. The brief template in the finishing cost guide works here too.
More in this topic: Problem solving: AI-built and no-code apps (26 guides) →