What a vibe-code rescue costs in the UK
It depends on how many problems there are and whether the data model is sound. UK contract developers advertised with React or Next.js skills had a median of about £516 to £529 a day in the six months to 10 November 2026, so a week of fixes at the median is roughly £2,600. Fixed-price rescues are priced after an audit instead.
On this page
What you are paying for
A rescue has two parts: finding out what is wrong, and fixing it. Most of the cost uncertainty comes from the first part. Until someone has looked at the code, the database and the accounts, nobody can tell you honestly whether you need three days or three weeks.
That is why quotes for fixing AI-built apps vary so much, and why a quote given without looking at the app is a guess. The work itself is ordinary software work: moving keys to the server, writing access rules, wiring up payment webhooks, adding error handling, fixing layouts. What makes it hard to price is that an AI-built app can look finished while missing most of those things.
UK day rates
The most useful public benchmark for UK developer rates is ITJobsWatch, which tracks the rates quoted in UK contract job adverts. These are the figures for the six months to 10 November 2026:
| Skill searched on ITJobsWatch | 25th percentile | Median | 75th percentile |
|---|---|---|---|
| Software Developer | £448 | £529 | £600 |
| React Developer | £425 | £516 | £650 |
| Next.js | £450 | £525 | £625 |
Two caveats. These are advertised contract rates, often for longer roles inside larger companies, not quotes for a short rescue. And the spread is wide: ITJobsWatch shows React Developer rates at £938 a day at the 90th percentile, against £381 at the 10th. Specialist, senior or short-notice work sits towards the top.
Day-rate arithmetic at those medians looks like this:
| Length of work | At £516 a day | At £529 a day |
|---|---|---|
| 3 days | £1,548 | £1,587 |
| 5 days (one week) | £2,580 | £2,645 |
| 10 days (two weeks) | £5,160 | £5,290 |
| 20 days (four weeks) | £10,320 | £10,580 |
These are not BlackFig prices. They are what the market median works out to if a developer bills by the day.
What pushes the cost up
Most rescues land on the same handful of problems. What changes the price is how deep each one goes.
| Smaller job when | Bigger job when | |
|---|---|---|
| Exposed keys | A few calls need moving to the server. | Most data access runs from the browser with a secret key. |
| Access rules | Tables have clear owners and only need policies. | Ownership is unclear, so the schema has to change first. |
| Data model | Small fixes: a missing constraint, a renamed column. | Duplicated data and text in place of links, with real user data to migrate. |
| Payments | Checkout works and only the webhook is missing. | Subscriptions, refunds and failed renewals have no code at all. |
| Error handling | A consistent data layer that needs error checks added. | Each screen fetches data its own way. |
| Mobile layout | Shared layout components with a few bad classes. | Every screen styled separately, with fixed widths. |
| Access to code and accounts | Code in your repository, accounts in your name. | Code locked in a builder, accounts in someone else's name. |
The single biggest cost driver is the data model. Fixes on top of a sound database are local and quick. Fixes on top of a wrong one ripple through every screen. If that is your situation, read when to refactor vs rebuild a vibe-coded app before getting quotes.
Three ways to pay for it
Day rate. You pay for time. It is flexible and simple, and it suits open-ended work where nobody knows the scope. The risk sits with you: if the job takes twice as long, it costs twice as much. Agree a cap and a check-in after the first two or three days.
Fixed price after an audit. Someone looks at the app first, then quotes a fixed price for a defined list of fixes. You know the cost before work starts. The risk of overruns sits with the developer, which is why the audit comes first. This is how I price rescues.
Agency project. An agency scopes, quotes and staffs the work. You get more people and more process, which can help on large apps and slows things down on small ones. You pay for the people around the developers too, such as project management, which is worth it on some projects and overhead on others.
There is more on the trade-offs in fixed price vs day rate, and on the different kinds of help in agency vs freelancer vs AI.
Some typical shapes of work
These are common patterns, not specific clients or quotes.
Security and payments only. The app works and users like it, but keys are in the browser, access rules are missing and paid access comes from the success page. The fix list is short and focused. At day rates, this is the cheapest kind of rescue.
A full pass before launch. Security and payments, plus error handling, mobile layout and a few tests around sign-up and checkout. This is the most common shape, and it is what a one-week sprint is built around.
Foundations first. The database needs reshaping before anything else can be fixed properly, and real users' data has to move with it. This is the most expensive shape, and the one where a partial or full rebuild should be on the table.
The cost of not fixing it
Most of what a rescue fixes costs nothing until the day it costs a lot.
- A data leak. If personal data is exposed, the ICO says you must assess the risk to the people affected and, if a risk is likely, report it as soon as possible and where feasible within 72 hours. That is a bad week for a small team, whatever happens next.
- Lost revenue. Payments that succeed in Stripe but never upgrade the account turn into refunds, support emails and churn.
- Lost users. People who hit a blank screen or a broken phone layout rarely report it. They just leave.
- Slower changes. Every new feature on fragile code takes longer and breaks something else, so the cost compounds month after month.
None of this means every app needs a rescue before it has users. A prototype for ten friendly testers can carry a lot of rough edges. But once strangers are signing up or paying, the top of the vibe-code rescue checklist stops being optional.
The BlackFig Rescue Sprint
I price rescues as a fixed price, agreed after a technical assessment (from £750), once I have seen the app. The sprint covers an audit plus the fixes: auth, data, payments, errors and mobile, made solid enough for real users. Most take about a week.
I don't publish a single figure, because the honest answer depends on what the technical assessment finds. You get the price in writing before any work starts. If the honest answer is that the app needs rebuilding, I will say so, and scope it like an MVP Build, which starts from £12,000 for 4 to 6 weeks.
How to keep the cost down
- Run the free checks yourself first. The vibe-code health check and the rescue checklist turn a vague worry into a list. A list is cheaper to quote.
- Prepare a handover pack. Time spent finding code, keys and accounts is time you pay for. How to hand a vibe-coded app to a developer lists what to gather.
- Stop adding features. Each prompt adds code that someone then has to check or fix.
- Prioritise. Fix what loses data or money first. Layout polish and refactoring can follow once users are safe.
- Own your accounts. Moving a domain or a Stripe account out of someone else's name takes time and goodwill.
For the bigger picture of why AI-built apps break and what fixing them involves, see my guide to fixing a vibe-coded app. If you are still at the build stage, MVP cost in the UK covers what a first version costs, and a Rescue Sprint is there when you are ready to get this one fixed.
- ITJobsWatch: Software Developer contract rates, UK, checked 10 November 2026
- ITJobsWatch: React Developer contract rates, UK, checked 10 November 2026
- ITJobsWatch: Next.js contract rates, UK, checked 10 November 2026
- ICO: UK GDPR data breach reporting (DPA 2018), checked 10 November 2026
The vibe-code rescue checklist, as a PDF
About 40 checks across auth, data, payments, errors, mobile and secrets, on two pages.
Questions founders ask
How much does it cost to fix an AI-built app in the UK?
At the median UK contract rate for React and Next.js work, about £516 to £529 a day in the six months to 10 November 2026, a week of fixes is roughly £2,600 and two weeks roughly £5,200 to £5,300. The real figure depends on how much is broken and whether the database needs reshaping.
Is it cheaper to rebuild than to fix?
Sometimes. If the data model is wrong or the code can't be exported, a rebuild can cost less than fixing everything in place. If the foundations are sound, fixing is almost always cheaper.
What does BlackFig's Rescue Sprint cost?
It is a fixed price, agreed after a technical assessment (from £750), once I have seen the app. Most take about a week. I don't publish a single number because the scope varies too much from app to app.
Why do day rates vary so much?
Experience, location, specialism and contract length all move the rate. ITJobsWatch's figures come from advertised contract roles, so a short freelance rescue may be priced differently.
Ant
Product designer who builds MVPs, based in Nottingham. 10+ years designing digital products, and the only person at BlackFig. More about me