A workflow, not one technology
The term is used loosely. Here it describes a workflow: state an intention in natural language, let an AI system propose or implement software, inspect the result, and iterate. A managed builder can combine that workflow with hosting and data services. A coding environment can apply it to an existing repository. Neither definition implies that the user must never inspect code.
The useful question is not whether a product qualifies for a fashionable label. Ask whether it can complete your intended task and whether your team can maintain the result. A marketing site and a multi-tenant application have different risks even if both begin with a prompt.
How it differs from no-code and autocomplete
Traditional no-code tools often expose a constrained visual model of pages, records, and workflows. AI builders may generate editable application code from a conversational brief. Coding assistants can work alongside a developer inside a broader engineering process. Modern products overlap these categories, so inspect the actual workflow rather than assuming a label defines all capabilities.
Macaly, Lovable, Bolt, v0, and Replit document prompt-driven building. Cursor documents codebase search, editing, terminal execution, and browser tools. The distinction in this directory is the operating model and intended handoff, not a claim that one category is inherently more capable.
Where human judgment matters
An AI system can produce a plausible implementation of an unclear request. It cannot establish that your business rules were correct in the first place. Define the required behavior, identify sensitive data, and test the failure cases. Review authentication, authorization, payments, and third-party integrations before accepting real users.
For public content, verify facts, licenses, disclosures, accessibility, and crawlable HTML. A successful build is a technical milestone, not evidence that a claim is true or that a site will rank. Start small, retain a recoverable version, and assign someone to own the application after launch.
Source notes
Evidence you can follow.
Official documentation consulted 7 November 2026. Product descriptions are vendor-published. Checklists and fit recommendations are editorial; no cross-tool benchmark is claimed.
- www.macaly.com/docs/en/welcome/faq ↗
- docs.lovable.dev/introduction ↗
- support.bolt.new/get-started/intro-bolt ↗
- v0.app/docs ↗
- docs.replit.com/replitai/agent ↗
- cursor.com/docs/agent/overview ↗
Continue reading
Your next useful step.
- Vibe coding for beginners: the learning path
- Full-stack AI apps: architecture learning path
- Deploying AI-built apps: the launch learning path
- AI app builder costs: the buying learning path