Join our Newsletter — 33% off our NHI Course
Home› Guides› What Is Lovable? AI App Builder Features and…
Guide AI Platforms

What Is Lovable? AI App Builder Features and Pricing: A Dated Reference for Teams

PDF for newsletter subscribers · free to read online

← All guides
By NHI Mgmt Group, edited by Lalit Choda Updated 15 November 2026 17 min read
On this page

Lovable is a full-stack AI development platform from Lovable Labs, based in Stockholm, that builds, edits and deploys web applications from natural language prompts. Its plans, credit rules and feature list change every few weeks, so much of what is written about it online is already out of date. This page sets out what Lovable is, what each plan costs, how credits work, which AI models apps can call, which modes and features exist, and which admin and data controls sit on each plan. It is written for founders, product teams, developers and the people who approve tools for an organisation. Every fact comes from Lovable's own website, documentation, changelog, blog and privacy policy. Last checked: 2026-11-15.

This is an independent guide from NHI Mgmt Group, based on Lovable's public documentation as of November 2026. It is not sponsored or reviewed by Lovable. Product features change, so confirm settings against Lovable's current documentation.

Key takeaways

  • A prompt-to-app platform with real code. Lovable's documentation describes it as "a full-stack AI development platform for building, iterating on, and deploying web applications", and the generated code can be edited and synced to Git.
  • Four plans, priced by credits. As of 15 November 2026 the plans are Free, Pro (from $25 per month), Business (from $50 per month) and Enterprise (contact sales), with Pro and Business tiers running from 100 to 10,000 monthly credits.
  • Three chat modes. Chat mode discusses changes without touching code, Plan mode writes an editable plan for approval, and Build mode implements and verifies changes.
  • A built-in backend. Lovable Cloud provides a database, authentication, storage, realtime updates, edge functions and secrets, and the documentation says it "utilizes Supabase's open-source foundation".
  • Governance scales with the plan. SSO and the Security center start on Business; SCIM, audit logs, EU inference and organisations are Enterprise features; Business and Enterprise data is excluded from AI training by default.

What is Lovable?

Lovable turns a written description of an app into a working web application with a frontend, backend, database, authentication and integrations. You describe what you want in a project chat, Lovable writes and changes the code, and you preview the result and publish it to a live URL. The code is not locked inside a visual editor: it can be opened in the code editor, synced to a Git provider and downloaded.

The documentation organises the product around four core concepts. An account is the individual user. A workspace is the shared space for a team, with a pooled credit balance. A project is one app. Chats are conversations outside any project, used for thinking and research, and work that needs building is sent to a project.

Lovable began life as GPT Engineer. Its blog post of 3 December 2024, "Rebranding: GPT Engineer to Lovable", marks the change of name. The company responsible for personal data under its privacy policy is Lovable Labs Sweden AB, Regeringsgatan 25, Stockholm, and the policy names the Swedish Authority for Privacy Protection (IMY) as its lead supervisory authority.

Lovable is used beyond the web editor. The documentation lists desktop and mobile apps, a Slack integration, Telegram, a ChatGPT app, the Lovable API for automating workspaces and a Lovable MCP server that external AI clients can connect to.

Key facts at a glance

As of 15 November 2026, from Lovable's own pages.

FieldValueSource page
CompanyLovable Labs Sweden AB, Stockholm, SwedenPrivacy policy
Product typeFull-stack AI development platform for web applicationsDocs welcome page
Former nameGPT Engineer (rebrand announced 3 December 2024)Lovable blog
Latest funding announced$400M Series C at a $13.3B valuation (12 August 2026)Lovable blog
PlansFree, Pro, Business, EnterpriseSubscription plans docs
Entry paid pricePro $25 per month or $250 per year for 100 monthly creditsSubscription plans docs
Free allowance5 build credits per day, capped at 30 per month, plus monthly Cloud, AI and connector grantsPlans and credits docs
Chat modesChat, Plan, BuildChat mode docs
Built-in backendLovable Cloud, built on Supabase's open-source foundationCloud docs
Backend regionsAmericas, Europe, Asia PacificCloud docs
Git syncGitHub, GitLab, BitbucketDocs welcome page, Git docs
Model behind the building agentNot stated by LovableAI features docs
Training on customer contentOpt out per account on Free and Pro; excluded by default on Business and EnterpriseSubscription plans docs, privacy policy

Plans and pricing

Prices below are from Lovable's subscription plans documentation as of 15 November 2026, in US dollars. Annual billing gives two months free.

PlanPriceMonthly creditsPositioned for
FreeNo subscription5 build credits per day, capped at 30 per month, plus 20 Cloud and 4 AI grant credits per monthTrying Lovable and smaller projects
ProFrom $25 per month or $250 per year100 to 10,000Fast-moving teams building together
BusinessFrom $50 per month or $500 per year100 to 10,000Growing departments needing advanced controls
EnterpriseContact salesVolume-based, set by contractLarge organisations needing flexibility and governance

Selected credit tiers, monthly price (USD), as of 15 November 2026:

Monthly creditsPro monthlyPro annualBusiness monthlyBusiness annual
100$25$250$50$500
400$100$1,000$200$2,000
1,200$294$2,940$588$5,880
5,000$1,125$11,250$2,250$22,500
10,000$2,250$22,500$4,300$43,000

How credits work

  • Usage grants on top of plan credits: Free, Pro and Business workspaces get a daily chat allowance, 5 daily build credits, and monthly grants of 20 Cloud credits, 4 AI credits and 1 connector credit. On Free, daily build credits stop once 30 have been granted in a calendar month; Pro and Business have no monthly cap on them.
  • Enterprise is a commitment balance: Enterprise has no daily chat allowance and no monthly AI grant. Admins see the remaining credits as a number.
  • Rollover: monthly plan credits roll over while the subscription is active. They expire two months after issue on monthly plans, or one month after the annual billing period ends on annual plans. Usage grants do not roll over.
  • Top-ups: one-time and automatic top-ups are available on Pro and Business and expire 12 months after purchase.
  • One balance for building and running: building, hosting, running the backend and AI features in a published app all draw from the same credit balance. Backend and AI usage is measured as Run credits.
  • Typical costs: the pricing page gives examples: "Make the button gray" 0.50 credits, "Remove the footer" 0.90 credits, "Add authentication with sign up and login" 1.20 credits, and "Build me a landing page, use images" 1.70 credits.

What this means for budgeting: a published app keeps spending credits after the build is finished, because backend and AI calls draw on the same balance. Usage limits and alerts are available on all paid plans, and per-member monthly credit limits from Pro upwards, so set both before an app goes live.

Models available

Lovable separates two kinds of model. The building agent writes, edits and reasons about your code; Lovable's AI features documentation does not name the model or models behind it. The app models are the ones your published app can call through Lovable AI, and the documentation says these are "not the models Lovable uses to write, edit, or reason about your code".

App models by use case, as listed on 15 November 2026 (default first):

Use caseModels listed
Chat and assistantsGPT-6 Astra (default), Gemini 3.8 Flash, Gemini 4 Argon, GPT-6 Luna, Gemini 3.1 Flash Lite; newer apps also offer Claude Fable 5.1, Opus 5.5, Opus 5, Sonnet 5 and Haiku 4.5
ImagesGPT Image 2.5 Sunburst (default), GPT Image 2.5 Flare, Gemini 3.1 Flash Lite Image
VideoGemini Omni 1.1 Flash (default), Veo 3.1
MusicLyria 3 Clip Preview (default), Lyria 3 Pro Preview
Search and RAGGemini Embedding 2 (default), text-embedding-3-small
Classification, routing, scoringJev 1.13 (default)
Text-to-speechGemini 3.1 Flash TTS Preview (default), Gemini 2.5 Pro TTS
Speech-to-textGPT Transcribe (default), Gemini 3.5 Transcribe
Live voiceGPT-Live 1 (default)
  • Billing: AI calls from deployed apps are counted as AI gateway usage and use credits. The monthly AI grant is used first. Live voice is billed by connected time with a 15-second minimum per call, and cancelled requests may still be billed.
  • When credits run out: app AI requests return 402 Payment Required, so an app's AI features stop working until the balance is topped up.
  • Churn: the model list moves fast. The changelog records GPT-6.1 Terra added on 8 November 2026, GPT-6.1 Sol on 15 October 2026, and Gemini 2.5 chat models retired for app AI features on 22 October 2026, which is a reason to check the AI features page rather than rely on this table.

Features and modes

Chat, Plan and Build

ModeWhat it doesCredit use
ChatDiscusses the app and possible changes without changing code, files or data; cannot create an editable plan, publish, or generate images or videosPriced as a chat message, typically a fraction of a credit; covered first by the daily chat allowance, which resets at 00:00 UTC
PlanInvestigates the project and writes a structured plan in the Plan view for you to edit and approve before anything is builtOne credit per message, plus any subagent research
BuildImplements changes and verifies the outcomeUsage-based build credits, not covered by the chat allowance

The conversation carries across all three modes, so you can discuss, plan and then build in one thread. The changelog shows Plan mode can compare and restore earlier versions of a plan (9 September 2026).

Other features

  • Lovable Cloud: database, storage, authentication, realtime, edge functions, secrets, scheduled jobs, logs and usage views. Turning on any one of the database, user accounts or file storage provisions all three. Since 7 October 2026, Cloud database changes are recorded as Drizzle migrations under drizzle/migrations/.
  • Supabase: since 20 October 2026 Lovable can create a Supabase project for an app with no backend.
  • Drafts: try changes in a separate copy of a project (9 September 2026).
  • Project memory and skills: Lovable remembers preferences and decisions across conversations (30 October 2026), "Dreaming" keeps project memory current in the background (4 November 2026), and personal skills and instructions arrived on 10 November 2026.
  • Automations: recurring work run from Chats (11 November 2026).
  • Import: projects can be imported from other tools or existing websites (29 October 2026).
  • Connectors: integrations that extend apps, with a monthly connector grant on Free, Pro and Business.
  • Sign in with Lovable: an identity option for apps and websites (14 October 2026).
  • WebMCP: browser agents can use your app through WebMCP tools (13 October 2026).
  • Security scans: a Quick scan runs automatically on publish; a Deep scan runs on request (see Enterprise and admin controls).

Supported IDEs and platforms

Lovable is not an IDE plugin. The main surface is the Lovable web app, with a code editor inside each project. The documentation also lists desktop and mobile apps; the desktop changelog records version 1.9 (tabs removed, 15 October 2026).

SurfaceWhat it is for
Web appProjects, chats, preview, code editor, publishing
Desktop and mobile appsWorking with projects and chats outside the browser
Git providersTwo-way sync with GitHub (github.com on all plans; GitHub Enterprise Cloud with data residency and GitHub Enterprise Server on Enterprise), plus GitLab and Bitbucket
Lovable MCP serverAny OAuth-capable MCP client can connect (11 September 2026)
Lovable APIAutomating workspaces (16 September 2026)
Messaging and assistantsSlack, Telegram and a ChatGPT app

Git behaviour worth knowing: the Lovable GitHub app requests write access to contents, pull requests, workflows and administration (used to create repositories). Each connection creates a new repository, private by default; Git sync cannot link to an existing repository, and reconnecting after a disconnect creates a new one. Lovable edits and syncs one branch at a time, usually the default branch. Paid plans can also download the codebase from Project settings or the code editor.

Enterprise and admin controls

Feature availability by plan, from the subscription plans documentation as of 15 November 2026:

ControlFreeProBusinessEnterprise
Workspace-private projectsYesYesYesYes
Custom domainsNoYesYesYes
User roles and permissionsNoYesYesYes
Role-based accessNoNoYesYes
Per-member monthly credit limitsNoYesYesYes
Single sign-on (SSO)NoNoYesYes
SCIMNoNoNoYes
Audit logsNoNoNoYes
Data excluded from AI trainingOpt out per accountOpt out per accountBy defaultBy default
Data Processing AgreementNoNoYesYes, plus custom DPA on request
  • Security center: available to workspace admins and owners on Business and Enterprise.
  • Block publishing with critical issues: set by workspace admins or owners; new Enterprise workspaces start with it on.
  • Enterprise-only security features: scheduled scans, and sensitive data (PII) findings within Security insights.
  • Enterprise governance added in 2026: audit log export to a SIEM or cloud storage (15 October), workspace policies and the Action center (16 October), organisations grouping several workspaces (20 October), IdP group mappings in one place (26 October), and a policy requiring admin approval before Lovable backend services are turned on (7 November).
  • Integrations: Wiz and Aikido can be connected at workspace level; the security documentation notes Aikido is billed separately.

Data handling and privacy

  • Training: the privacy policy says Lovable may use Customer Content and Usage Data to train and improve its models, and does not train on account and billing details or on your app users' data. You "can opt out of model training at any time in your account settings, on any plan, at no cost". The opt-out applies going forward and does not withdraw content from datasets or models built earlier. For Business and Enterprise the prohibition is set out in the Data Processing Agreement.
  • Retention: account and customer content are kept while the account is open, then deleted or de-identified after closure or a verified deletion request. The policy gives no fixed periods; security logs are kept as long as needed, and billing records as tax law requires.
  • Model providers and sub-processors: the policy points to trust.lovable.dev for the sub-processor list with locations and purposes, and says its agreements with AI model providers restrict their use of customer content, including for training.
  • Location: data is processed in several countries including the United States, with EU Standard Contractual Clauses and the UK Addendum for transfers. Lovable Cloud projects run in Americas, Europe or Asia Pacific; the region is fixed once the backend is enabled, Business and Enterprise admins can set a workspace default, and country-specific regions are available to Enterprise on request. EU inference is listed as an Enterprise feature.
  • Compliance documents: the security documentation does not itself claim a certification; it says Aikido pentest reports are structured for SOC 2, ISO 27001 and security questionnaires. Check trust.lovable.dev for Lovable's current attestations.

For guidance on database permissions in Supabase-based backends, see the Supabase Security guide and the Supabase Keys Security Guide.

Release history

Major milestones from Lovable's blog and changelog.

DateMilestone
3 December 2024GPT Engineer rebrands as Lovable
28 January 2025Blog: "Zero to $10M ARR in 2 months"
25 February 2025$15M additional funding announced
24 April 2025Lovable 2.0
30 June 2025Agent Mode (Beta) introduced
17 July 2025$200M raised at a $1.8B valuation
23 July 2025Blog: "$100M ARR and Lovable Agent"
29 September 2025Lovable Cloud and Lovable AI introduced
18 December 2025$330M raised
12 August 2026$400M Series C at a $13.3B valuation, led by Menlo Ventures and the Scaleup Europe Fund managed by EQT
16 September 2026Lovable API
21 September 2026Chats outside projects and Chat mode
15 to 20 October 2026Audit log export, workspace policies, Action center, Enterprise organisations
11 November 2026Automations from Chats

Limits and known constraints

  • Free build cap: 5 build credits per day, stopping at 30 in a calendar month until the 1st of the next month.
  • Credit expiry: plan credits expire two months after issue on monthly billing; top-ups after 12 months; daily and monthly grants do not roll over.
  • Running apps cost credits: backend and AI usage of a published app draw from the workspace balance, and app AI calls fail with HTTP 402 when it is empty.
  • Region is permanent: a Lovable Cloud project's region cannot be changed after the backend is enabled.
  • Git constraints: new repository per connection, no linking to an existing repository, one active branch at a time.
  • Chat mode boundaries: Chat mode cannot publish, edit code or create an editable plan.
  • Enterprise allowances: not published; set by contract.

Security guidance

Apps built in Lovable are your apps: access rules, secrets and what you publish are customer-side decisions. The Lovable-specific controls are covered in depth in the Lovable Security Guide; the points below are the ones that follow from the facts on this page.

  • Check per-record access before publishing: the Quick scan's database review flags tables without per-record access control. Review policies under More, Database, Manage, Policies, and run a Deep scan before a public launch.
  • Turn on publish blocking: set "Block publishing with critical issues" at workspace level (on by default for new Enterprise workspaces).
  • Keep secrets in Cloud secrets: store API keys in the backend's secrets rather than in prompts or frontend code, and rotate the Cloud database password from chat when people leave a project.
  • Treat the GitHub app as a privileged integration: it holds write access to contents, workflows and administration. Install it on a dedicated organisation or limited repositories and review it with your other GitHub apps.
  • Govern MCP and API access: the Lovable MCP server accepts any OAuth-capable client and the Lovable API automates workspaces. Track which clients and tokens have access, with an owner for each.
  • Use leaked password protection: the security documentation describes a Have I Been Pwned check for email and password authentication in apps.

For the broader risks of building software by prompt, see the Vibe Coding Security Guide and the Top 10 Vibe Coding Security Risks.

Practitioner checklist

  • Record the date and credit tier whenever you quote a Lovable price, and recheck the subscription plans page before renewal.
  • Use Business or Enterprise for organisational work so SSO, role-based access and default training exclusion apply, and sign the DPA.
  • On Free or Pro, turn on the model training opt-out in account settings for every member.
  • Set usage limits, alerts and per-member credit limits before publishing any app that calls backend or AI features.
  • Choose the Lovable Cloud region deliberately before enabling the backend, since it cannot be changed later.
  • Run a Deep scan and review database policies before every public launch, and turn on publish blocking for critical issues.
  • Inventory the Lovable GitHub app installation, MCP clients and API tokens, each with a named owner and review date.
  • Re-read the Lovable changelog monthly, as models, credit rules and admin controls change frequently.

Standards and references

Related NHI Mgmt Group resources: Lovable Security Guide · Vibe Coding Explained · Vibe Coding Security Guide · Supabase Security · AI Coding Agents Security Guide

PDF for subscribers

Take this guide with you

Every section, table and checklist in one PDF, with links to the sources and related guides. Subscribe to our weekly NHI & AI Identity Journal and we’ll email you a download link.

Free weekly newsletter

Subscribe to the NHI & AI Identity Journal

The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.

Bonus 33% off our NHI Course when you subscribe.

NHIMG Editorial Note
Researched and written by the NHI Mgmt Group editorial team using AI tools, under the editorial direction of Lalit Choda. Facts are drawn from the sources listed on this page and checked by our automated quality process. Last updated 15 November 2026. Spotted an error? Use our contact page.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM and Agentic AI security. nhimg.org

PDF Downloads for Newsletter Subscribers

Get the guide

Enter your details and we’ll email you a download link.

The full guide is free to read online. The PDF is a thank you for subscribing.

Fill in all fields marked * and tick the PDF box to continue.