Updated 21 Oct 2026 · measured 16 Oct 2026 · test v1 · 1 run per tool · free plans · single observation
What one real app actually cost on each free plan (measured, not list prices)
Price trackers log list prices. We logged what one real app actually cost on Lovable, Bolt, Base44 and v0's free plans: the same prompt on each, usage read off each tool's own usage page before and after, with screenshots.
Short answer · as of 21 Oct 2026
One working multi-user app with login and a database used 2.10 credits on Lovable, about 101K tokens on Bolt, 1 message credit on Base44 and $3.81 on v0 Observed. That's 7% of Lovable's free month, 10% of Bolt's, 4% of Base44's and 76% of v0's one-time $5 Inferred. All four were live with zero fix prompts, in about 3.5 to 15 minutes Observed.
Measured 16 Oct 2026, one run per tool: a single observation, not an average. Status: Current as of 16 Oct 2026 for all four (goes stale 14 Jan 2027 under our 90-day rule; last checked for major updates and price changes 6 Nov 2026). Details · how status works
Same prompt, fresh free account per tool, default settings, no top-ups, 16 Oct 2026. Usage is what each tool's own usage or billing page showed before and after the build (screenshots in each tool's section below). Percentages are rounded to the nearest whole number. Dollar values are usage × the list rate of the cheapest paid plan, checked 16 Oct 2026; you'd pay the subscription, not this amount. v0 also needed six manual setup steps (Vercel terms, Neon install, two SQL approvals, an auth secret, a manual Publish); the other three set up login and the database themselves.
How fast do free credits run out?
Short answer · measured 16 Oct 2026 · single observation
One clean build of our app used 42% of Lovable's daily build credits, 34% of Bolt's daily tokens, 20% of Base44's daily credits and 76% of v0's one-time $5 Observed usage, Inferred %. So the daily allowance runs out after about 2 builds on Lovable and Bolt and 5 on Base44, and v0's free credit after 1 build in total. Iterating costs more than a first build, and we didn't measure fixes.
Tool
One clean build used Observed
Same-size builds before the daily cap Inferred
Same-size builds in a free month Inferred
When it comes back Verified
Cheapest paid plan, same-size builds a month Inferred
Plus, $30 of credit: about 7, plus $2 on each daily login
Builds that fit = each vendor's allowance ÷ our one measured build, rounded down. That assumes every build costs what our first, fix-free build did, which real projects won't, so treat these as the most you'd get, not what you'll get. Paid-plan counts assume paid credits meter like free ones; Lovable says credit value can differ between plans. Allowances Verified on vendor pages read 18 Oct 2026 and re-read (Lovable, Bolt) 21 Oct 2026.
Why credits can go much faster than this
Builders report burning through credits far faster than our numbers suggest: one post on X, seen in our 21 Oct 2026 market watch, said 100 Lovable credits went in under an hour Reported, single post, not tested by us. Both can be true. Our run was one prompt and zero fix prompts; the vendors' own docs say what makes usage grow from there:
Lovable: Build mode cost "depends on the complexity of the request and the work completed". Lovable's own examples run from 0.50 credits ("make the button gray") to 2.00 (a landing page with images); a Plan mode message is 1 credit plus any research. A single long-running message can run for up to 10 hours and only checks in with you at 20 credits by default, and /goal messages cost "noticeably more". Verified, docs.lovable.dev credits and usage, read 21 Oct 2026
Bolt: "Most of your token usage comes from Bolt reading, understanding, and syncing your project files, so larger projects use more tokens per message." Direct edits in Code view use no tokens. Verified, support.bolt.new tokens page, read 21 Oct 2026
Base44: On Free, automatic error fixes cost message credits; stopping the AI partway now charges for the replies it already finished. Manual visual edits use no credits and a Discuss mode message uses about 0.3. Verified, Base44 docs and changelog, read 18 and 19 Oct 2026
v0: Credits are dollars and generation pauses at zero; on a new account there is no way to buy more without upgrading. Our one build used $3.81 of the $5, leaving $1.19 for fixes. Verified, v0 docs, read 18 Oct 2026; usage Observed 16 Oct 2026
Our reading Inferred: the first build is the cheap part. Long agent runs, edits on a growing project and fix loops are where a free allowance goes. To stretch it, plan before you prompt (Lovable's Chat mode has its own daily allowance; Base44's Discuss mode costs about 0.3 credits), make small edits by hand where the tool allows it for free (Bolt's Code view, Base44's visual edits), and check what each message cost (Lovable shows it under each reply). Our v2 test is a bigger app run 3 times per tool and reports the free allowance each run used; those numbers are on the v2 results page (published 25 Oct 2026).
What we didn't record
We only publish numbers that are in our run receipts. These weren't recorded, so they aren't estimated here either:
Bolt's exact token count. Bolt shows balances rounded to the nearest 1K, so "about 101K" is as precise as the receipt gets.
Backend and hosting use after launch. Lovable Cloud credits, v0's Neon database and Vercel hosting, and Bolt's web requests weren't recorded. For Base44 they were: integration credits stayed at 0 of 100.
v0's daily message count. Its free plan also limits you to 7 messages a day; how many of those the build used wasn't recorded.
The cost of fixing or extending an app. All four needed zero fix prompts, so we have no data on what a fix costs. Bigger apps and fixes will cost more.
Paid-plan usage. Every run was on a free plan. A paid plan may meter the same build differently (see caveats).
What list-price trackers can't tell you
Several sites now log list-price changes for AI app builders Observed, 20 Oct 2026. That's useful: it tells you when a plan or an allowance moved, and we keep dated price pages too (all four). But a list price and a measured build answer different questions:
Question
From list prices
From a measured build (this page)
What does each plan include, and when did it last change?
Yes. This is what price logs are built for.
Only on the day we tested. We re-read prices separately on our pricing pages.
How much of a free plan does one real app use?
No. Vendors price in credits, tokens or dollars and don't say how many one app takes.
Yes. We read each tool's own usage or billing page before and after the build.
Did the app actually work, and how many fix prompts did it need?
No.
Yes. 10 feature checks, a second-user data check and a phone-size check; 0 fix prompts on all four.
How long from prompt to a live URL?
No.
Yes. About 3.5 to 15 minutes in our runs.
What is one app worth at paid rates?
Only if you guess the usage.
Measured usage × the list rate (Inferred, and labelled that way).
Is the number typical?
Not applicable.
Not yet: one run per tool, one fairly simple app. That is a limit of our data.
The two fit together: the list price gives you the rate, and a measured build tells you how much you'll use at that rate. For each tool's full plan list, yearly prices and what happens when the free plan runs out: Lovable pricing · Bolt pricing · Base44 pricing · v0 pricing.
How much of the free plan one app used
Base444% of the 25 a month
Lovable7% of the 30 a month
Bolt10% of the 1M a month
v076% of the one-time $5
Monthly allowance for Lovable, Bolt and Base44; v0's free credit is one-time, which is why its bar is so long. Inferred from our Observed usage and each vendor's Reported allowance.
What one app is worth at paid-plan rates
Base44≈ $0.16 to $0.20
Bolt≈ $0.25
Lovable≈ $0.53
v0≈ $3.81
Usage × the per-unit rate of the cheapest paid plan open to new users, month-to-month list price (Base44 shown as a range because its page leads with the yearly price), 16 Oct 2026. Inferred. You'd pay the subscription, not this amount.
A one-time $5 credit for new sign-ups, plus a limit of 7 messages a day
1, ever
Plus, $30 of credit a month: $30 per user a month
"Builds on the free plan" assumes every build costs the same as our single run, which real builds won't.
Are these numbers current?
These costs come from the same v1 runs as our reviews, so they share their status. A result is current for 90 days from its test date, then stale, and it needs a retest sooner if a tool ships a major update or a pricing change that affects the test (rules).
Retest coverage: 4 of 4 published results are current (100%). 0 stale, 0 need a retest. v2 results published for 3 of 4 (v2 results). The next result goes stale on 14 Jan 2027 unless it's retested first. (Status computed when this page was built, 6 Nov 2026.)
v2 is a harder app, run 3 times per tool, and its pre-registered plan reports cost per run separately from the score. The v2 results (published 25 Oct 2026) list the free allowance each run used; on the free plans, 3 of the 9 runs ran out of free credits before the app deployed (Lovable run 3, Bolt runs 2 and 3). v0 is not in v2 (decided 19 Oct 2026): its one-time free credit was mostly used in v1 ($3.81 of $5) and we don't buy top-ups during a test, so the v0 figure above stays its only measured cost. Get the next test first.
Built an app on one of these plans? Our $29 Launch Check runs real-browser checks on your live app (desktop and phone, sign-up as a new user, up to 25 pages) and sends a private report with fix prompts within 48 hours. It's separate from our rankings: your app is never ranked, and payment never changes a tool's score or tier.
Each tool's own billing or usage page, before and after the run (screenshots below)
Time to live
From sending the prompt to a working public URL, from the run's timestamped log
Pricing checked
16 Oct 2026, on each vendor's official pricing page and docs (links below); re-read 18 Oct 2026 on our pricing pages
Current as of
16 Oct 2026 (status: Current; goes stale 14 Jan 2027)
Page updated
21 Oct 2026: "How fast do free credits run out?" section and 2 FAQs added (Lovable and Bolt docs re-read 21 Oct 2026); 20 Oct 2026: measured-cost summary, per-tool table, "what we didn't record" and method comparison added. No measured numbers changed
Caveats (read before citing)
One run per tool. Usage varies between runs, so treat each number as one data point, not an average. Our v2 test runs 3 times per tool and will report a median and range.
v1 was easy. All four tools scored 95 to 100 and needed zero fixes, so this is the cost of a simple app built cleanly. Harder apps, and apps that need fix prompts, will cost more.
Units differ. Credits, tokens and dollars aren't the same thing. The dollar column turns them into one unit using list prices; it is not what you'd be billed.
Paid plans may meter differently. Lovable says credits are "not necessarily equal in value across different plans", Base44 says usage varies by action and model, and v0 lists "access to all models" only on its paid plans. A paid-plan build could use a different amount than our free-plan build.
Included grants can make it $0. Lovable Pro includes 5 daily build credits and v0 Plus gives $2 of credit on each daily login, so on those plans a build this size may not touch paid credits at all.
Yearly billing lowers the rate. The dollar column uses month-to-month prices for Lovable, Bolt and v0. Lovable and Bolt also sell Pro billed yearly, which makes each credit or token cheaper; yearly prices are on the Lovable and Bolt pricing pages.
v0's free credit changed during our run. Our billing page said "$5 in credit every month" before the run and "a one-time $5 credit. This credit does not renew" after it (Observed, screenshots below). v0's docs say new sign-ups get $5 once and existing Free users keep $5 a month, so your account may differ.
Prices change. Vendor pricing is Reported as of 16 Oct 2026; we re-read every price on 18 Oct 2026 and none had changed (pricing pages). Check the linked pages before you buy.
Prompt to live appObserved: under 4 min, with 0 fix prompts. Setup: None. Lovable Cloud set up auth and the database automatically.
Free allowanceReported: 5 build credits a day, capped at 30 a month.
Share usedInferred: 42% of the daily 5; 7% of the 30 a month.
Same-size builds on the free planInferred: 2 a day, about 12 a month. 5 ÷ 2.10 = 2 builds a day. Free grants stop after 30 credits (6 days), so 6 × 2 = 12.
Cheapest paid planReported: Pro, 100 credits a month, $25 a month ($250 a year) ($0.25 per credit).
Value of this build at that rateInferred: ≈ $0.53. 2.10 × $0.25. Pro also includes 5 daily build credits, which would cover a build this size before any paid credits are used.
Status: Current, as of 16 Oct 2026 · 1 run · Single observation. What the app looked like and what broke: Lovable review.
All plans and free-plan limits, re-read 18 Oct 2026:Lovable pricing
UsedObserved: 1 message credit. Credit usage page: 0 of 25 → 1 of 25 message credits; integration credits 0 of 100, unchanged.
Prompt to live appObserved: about 3.5 min, with 0 fix prompts. Setup: None. Built-in database + auth. Free plan requires the app to be public.
Free allowanceReported: 25 message credits a month, up to 5 a day.
Share usedInferred: 20% of the daily 5; 4% of the 25 a month.
Same-size builds on the free planInferred: 5 a day, about 25 a month. 5 ÷ 1 = 5 builds a day; 25 ÷ 1 = 25 builds a month.
Cheapest paid planReported: Starter, 100 message credits a month, $16 a month billed yearly; about $20 month to month (Inferred, see below) ($0.16 to $0.20 per credit).
Value of this build at that rateInferred: ≈ $0.16 to $0.20. 1 × ($16 or $20) ÷ 100. Base44 lists $16/month billed yearly and says yearly saves 20%; its site also says paid plans start "from $20/mo", which matches about $20 month to month.
Status: Current, as of 16 Oct 2026 · 1 run · Single observation. What the app looked like and what broke: Base44 review.
All plans and free-plan limits, re-read 18 Oct 2026:Base44 pricing
Base44 credits before: 0 of 25 usedBase44 credits after: 1 of 25 used
v0: $3.81 of credit
UsedObserved: $3.81 of credit. Billing page: $5.00 → $1.19 available.
Prompt to live appObserved: about 15 min, with 0 fix prompts. Setup: Six steps: accept Vercel terms, install Neon, two SQL approvals, generate an auth secret, manual Publish.
Free allowanceReported: A one-time $5 credit for new sign-ups, plus a limit of 7 messages a day.
Share usedInferred: 76% of the one-time $5.
Same-size builds on the free planInferred: 1, ever. $5 ÷ $3.81 = 1 build. For new accounts the $5 is one-time and does not renew.
Cheapest paid planReported: Plus, $30 of credit a month, $30 per user a month (credits are in dollars).
Value of this build at that rateInferred: ≈ $3.81. Face value: $3.81 is about 13% of Plus's $30 monthly credit. Plus also gives $2 of free credit on each daily login, which could cover part of it. The $20 Premium plan is being retired and is closed to new users.
Status: Current, as of 16 Oct 2026 · 1 run · Single observation. What the app looked like and what broke: v0 review.
All plans and free-plan limits, re-read 18 Oct 2026:v0 pricing
VibeTierList, "What one real app actually cost on each free plan: Lovable, Bolt, Base44 and v0 (measured)", prompt-to-app test v1, 1 run per tool (single observation), tested 16 Oct 2026, page updated 21 Oct 2026. https://vibetierlist.com/research/cost-per-app/
Evidence labels:Observed = we saw it in our own run (screenshots above). Reported = the vendor says so on its own pages. Inferred = our arithmetic from Observed and Reported numbers. Single observation = seen in one run. What the labels mean. Disclosures: no tool on this page paid us, and none of these links are affiliate links. Payment never buys rank. Spot an error? [email protected].
FAQ
How much does it actually cost to build an app on Lovable, Bolt, Base44 or v0?
Measured on 16 Oct 2026, one run per tool (single observation): one real working app (a multi-user to-do app with login, a database and a public URL, from the same prompt) cost 2.10 build credits on Lovable's free plan, about 101K tokens on Bolt's, 1 message credit on Base44's and $3.81 of v0's free credit. That is 7% of Lovable's free month, 10% of Bolt's, 4% of Base44's and 76% of v0's one-time $5. All four got to a live app with zero fix prompts, in about 3.5 to 15 minutes. At each tool's cheapest paid-plan rate (list prices checked 16 Oct 2026) that usage is worth about $0.53 on Lovable, $0.25 on Bolt, $0.16 to $0.20 on Base44 and $3.81 on v0 (Inferred). On the free plans themselves nothing was charged.
How fast do free credits run out on Lovable, Bolt, Base44 and v0?
On the free plans, one clean build of our simple app (one prompt, zero fixes) used 42% of Lovable's daily build credits, 34% of Bolt's daily tokens, 20% of Base44's daily credits and 76% of v0's one-time $5 (measured 16 Oct 2026, single observation). So Lovable's and Bolt's daily allowance runs out after about 2 builds that size, Base44's after 5, and v0's free credit after 1 build in total. Lovable's whole free month (30 credits) is granted in the first 6 days. Iterating costs more than a first build: Lovable and Bolt both say larger requests and larger projects use more per message, and we did not measure fix prompts.
Why do some people say 100 Lovable credits go in under an hour?
We saw that claim in a single post on X during our 21 Oct 2026 market watch (Reported; we have not tested it). It doesn't contradict our number. Our clean build used 2.10 credits from one prompt with zero fixes, so 100 credits would cover about 47 builds that size (Inferred). Lovable's docs say Build mode cost depends on the complexity of the request and the work done, a single long-running message only checks in at 20 credits by default, and /goal messages cost noticeably more (Verified, read 21 Oct 2026). Long agent runs, big edits and fix loops are where credits go, and our v1 data has no fix prompts to measure.
Why measure usage instead of comparing list prices?
A list price tells you the rate: how many credits, tokens or dollars a plan includes. It doesn't tell you how many of them one app takes, because none of the four vendors publishes a usage figure for a whole app. We built the same app on each free plan and read each tool's own usage page before and after, so you can see how far a plan goes. List-price logs are still the better way to see when a plan changes; the two answer different questions.
Which AI app builder gives you the most free builds?
Base44, on our numbers. One build used 1 of its 25 monthly message credits, so the free plan fits about 25 builds that size a month. Lovable fits about 12, Bolt about 9, and v0 one in total, because new accounts get a one-time $5 credit (Inferred from one run each; real builds vary).
Can you build a full app on v0's free plan?
Once. Our build used $3.81 of the $5 free credit and finished with no payment. For new accounts that credit is one-time and does not renew, so a second build of the same size would not fit (Observed on our billing page; v0 docs say existing Free users may still get $5 a month).
What is the cheapest paid plan for each tool?
As of 16 Oct 2026 (Reported), unchanged when we re-read them on 18 Oct 2026: Base44 Starter at $16 a month billed yearly (about $20 month to month), Lovable Pro at $25 a month for 100 credits, Bolt Pro at $25 a month for 10M tokens, and v0 Plus at $30 per user a month. You pay the subscription, not per app. The per-app dollar figures here are the value of the usage at those rates. Every plan, with yearly prices, is on our pricing page for each tool.
How current are these numbers?
Current: the results these costs come from were tested on 16 Oct 2026 and go stale on 14 Jan 2027 under our 90-day rule, sooner if a tool ships a major update or a pricing change that affects the test. We last checked for those on 6 Nov 2026. Each figure is a single observation: one run per tool.
Is there cost data from the harder v2 test?
Yes. The v2 results (published 25 Oct 2026, vibetierlist.com/results/prompt-to-app-v2/) list the free allowance each of the 9 runs used, separately from the score. On the free plans, 3 of the 9 runs ran out of free credits before the app deployed (Lovable run 3, Bolt runs 2 and 3). This page still shows v1 cost only. v0 is not in v2 (decided 19 Oct 2026): its one-time free credit was mostly used in v1 ($3.81 of $5) and we don't buy top-ups during a test, so the v0 figure on this page stays its only measured cost.