Method

How we test AI app builders

Every tool in the tracker is assessed on the same criteria, against the vendor's own pricing page, with a last-checked date. This page explains what we look at and how we score it.

No paid rankings. Some links are affiliate links; they never change our scores. Disclosure

Criteria

What we assess

Price from. The lowest paid monthly plan, in USD, billed monthly. Annual rates, per-user pricing and usage or credit billing are noted separately, because they change what you actually pay.
Free plan. Whether one exists and its limits: credits or messages, publishing, branding, custom domains and code export.
Non-coder difficulty. How far someone who has never written code can get without help. Scored 1-5 on the scale below.
Data location. Where your app and its data are hosted, whether a database and login are built in, and whether you can take the code with you.
Data and security. Training settings, single sign-on and roles, who owns the platform, and billing that can exceed the headline price.
Best for. A one-line verdict on the job each tool suits best.

Scoring

The 1-5 non-coder difficulty scale

The test is the same for every tool: can someone with no coding background get from a blank prompt to a working, shareable tool with login and stored data? Higher scores mean easier.

5 Very easy: Realistic in one sitting, with no help. Currently: Lovable, Base44
4 Easy: A non-coder gets to a working tool, but with more steering and more time than the 5/5 tools. Currently: Replit, Bolt.new
3 Moderate: Real progress without code, but expect trial and error or some technical help to reach a finished, shared tool. Currently: v0, Make
2 Hard: Built for people who already work with code. You will need a technical partner or real time spent learning. Currently: Cursor, Windsurf
1 Very hard: Not realistic without writing or reading code yourself. Currently: None at present

Sources

How we check sources

Prices and plan limits come from the vendor's own pricing page first. We record the date we checked it.
We compare each tool against one independent secondary source.
Where sources disagree, the vendor's current page wins for price, and we flag the difference as "not fully verified" in the tracker and on the tool's review. We also flag anything we could not confirm, such as free-plan limits the vendor does not publish.
When we get something wrong, we fix it and update the last-checked date. Spotted a change? [email protected].

Sources, open flags and last-checked dates

Primary source, secondary source, last-checked date and open flags for each tool
Tool	Vendor source	Secondary source	Last checked	Open flags
Lovable	lovable.dev	getolune.com	2026-10-16	Free tier limits change often
Replit	replit.com	getolune.com	2026-10-16	Free Starter plan is not shown on replit.com/pricing; documented at https://docs.replit.com/billing/plans/starter-plan
Bolt.new	bolt.new	getolune.com	2026-10-16	None flagged
Base44	base44.com	getolune.com	2026-10-16	None flagged
Cursor	cursor.com	flowrushtech.com	2026-10-16	Annual pricing not shown on official page Plans now sold via X Subscription Corp bundle (Grok); model access differs by tier
v0	v0.app	getolune.com	2026-10-16	Free tier message limits
Windsurf	devin.ai	flowrushtech.com	2026-10-16	windsurf.com/pricing now redirects to devin.ai/pricing (rebrand to Devin)
Make	make.com	n8n.io	2026-10-16	Paid prices scale with monthly credit volume; figures shown are the 10,000-credit tier Some AI features use more than 1 credit per action

Data from the tracker, last checked 2026-10-16.

Independence

Affiliate links and scores

No paid rankings. Some links are affiliate links; they never change our scores. Scores come from the criteria above, and tools are listed whether or not they have an affiliate programme. Full details are in our disclosure.

See the tracker