# Lovable AI platform that builds and publishes web apps with frontend, backend and database from a description. Canonical profile: https://ki-agentenkompass.de/en/agents/lovable/ Editor: [Michael Meierhoff](https://ki-agentenkompass.de/en/about/) Sources reviewed: 2026-11-05 Basis: provider documentation. Examples are editorial suggestions, not hands-on product tests. ## Overview Lovable is a hosted product from Lovable Labs Incorporated. The reviewed pages named no public source-code licence. - Provider: Lovable Labs Incorporated - Product family: Lovable - Profile type: AI agent platform for building web apps - Environment: Browser; also a desktop app, a mobile app, Slack and an MCP server - Connections: Git sync (GitHub, GitLab, Bitbucket), connectors such as Slack, Notion or HubSpot, built-in backend (Lovable Cloud) or Supabase - Setup: Create an account, describe your app, check the preview and publish. - Pricing model: Free plan with daily limits; Pro from $25 a month, Business from $50. Everything draws on one credit balance. ## Key features ### Build and Plan modes In Build mode Lovable implements changes directly and, per the documentation, can work on one message for up to 10 hours. Plan mode writes a plan you edit and approve before any code is changed. Source: https://docs.lovable.dev/features/agent-mode ### Instructions, skills and memory Project and workspace instructions apply in every conversation, and skills are reusable playbooks. Lovable also saves what it learns while you build as project or personal memory. Source: https://docs.lovable.dev/glossary ### Git sync A project can be linked to a repository you own on GitHub, GitLab or Bitbucket. Changes flow in both directions, but you cannot import an existing repository. Source: https://docs.lovable.dev/integrations/git-sync-overview ### Backend and connectors Lovable Cloud includes a database, authentication, storage and edge functions; alternatively you connect your own Supabase project. Connectors link tools such as Slack, Notion or HubSpot. Source: https://docs.lovable.dev/integrations/introduction ### Scheduled jobs Jobs are scheduled background tasks in the backend, for example every day at 08:00. You create them through chat or SQL; the Jobs view shows the schedule and run history. Source: https://docs.lovable.dev/features/jobs ### Security checks A quick scan runs automatically when you publish, and the deep scan is started manually. Wiz and Aikido pentests are optional additions. Source: https://docs.lovable.dev/features/security ### Lovable in Slack Mention @Lovable in a channel or direct message to create projects or change existing ones. Lovable can only see projects the requesting person has access to. Source: https://docs.lovable.dev/integrations/lovable-for-slack.md ## Example workflows ### First web app from a description - Your input: A plain-language description of the app, ideally without logins and a database at first. - Possible result: Lovable builds a first version in the preview, which you change step by step and publish at a lovable.app address. - What to check: Check buttons, links and the mobile view in the preview before publishing. ### Scheduled task in the backend - Your input: An app with Lovable Cloud and a description of a recurring task, such as a daily summary. - Possible result: Lovable creates a job with a schedule that you switch on and off in the Jobs view. - What to check: Check the run history in the Jobs view and credit usage regularly. ## Requirements and limits ### Account needed Requirement: You need an account to build and publish. Sign-up works with email, Google, GitHub or Apple. Source: https://docs.lovable.dev/introduction/getting-started.md ### Free plan restrictions Limitation: Free cannot edit or download code. Publishing to invited people only is limited to 10 people per project, and daily build credits are capped at 30 per calendar month. Source: https://docs.lovable.dev/introduction/subscription-plans.md ### Work stops without credits Limitation: When no credits are left, building stops. AI features in deployed apps stop working, and apps using the built-in backend pause. Source: https://docs.lovable.dev/introduction/credits-and-usage.md ### Git sync is export only Limitation: You cannot import an existing repository into Lovable. Each project has one repository and one synced branch. Source: https://docs.lovable.dev/integrations/git-sync-overview ### Security stays your responsibility Limitation: Per the documentation you are responsible for ensuring your app meets the security requirements for its use case. The built-in scans are static analysis, an optional Aikido connection adds dynamic testing, and the tools cannot guarantee complete security. Source: https://docs.lovable.dev/features/security ### Not HIPAA-compliant Limitation: By its own statement, Lovable does not currently sign Business Associate Agreements and is not HIPAA-compliant. Source: https://docs.lovable.dev/introduction/lovable-for-enterprise.md ## A good fit if … - you want a published web app with a backend from a description without your own setup - you want to use Git sync, Slack or connectors and keep instructions for all projects - you want to set up scheduled background tasks directly in your app's backend ## Less suitable if … - you want to continue an existing repository in Lovable, as Git sync only exports - you need HIPAA compliance for health data - you want predictable costs without credit consumption Editorial assessment based on provider documentation. Not a product test. ## Usable AI models Fixed choice across providers Lovable chooses the model for Build, Plan and Chat itself, and you cannot change it. Only the EU inference page for Enterprise gives names: Claude and GPT for text, Gemini only for images and embeddings. The model choice for AI features inside your app is separate and does not apply to the agent. - Model families named: Anthropic (Claude), OpenAI (GPT) - Model choice: The product chooses - Your own API key: Not stated - Router or custom endpoint: Not stated - Local models: Not stated Partly documented · Sources accessed on: 2026-11-05 - [Lovable: FAQ](https://docs.lovable.dev/introduction/faq) - [Lovable: EU inference](https://docs.lovable.dev/features/eu-inference.md) - [Lovable: AI features for your app](https://docs.lovable.dev/features/ai) - [Lovable: The model picker is a dead end](https://lovable.dev/blog/the-model-picker-is-a-dead-end) ## Europe & regional availability Documentation and terms reviewed. EU inference is described explicitly; the pages state no general country release for Europe. Reviewed scope: Hosted service; AI processing and data storage treated separately Reviewed on: 2026-11-05 ### EU inference on Enterprise only With requirements As a workspace setting for Enterprise, Lovable sends AI requests to model endpoints in the EU. If no EU endpoint can serve a request, it fails. It uses more credits and some features are unavailable, such as video, voice and Dreaming. Region covered: EU, parts of the EEA; United Kingdom excluded Documented reason: The page names the setting for Enterprise only and says it covers where AI requests are processed, not where published apps store data. Web search, web fetch and browser automation process data in the United States. Official source: https://docs.lovable.dev/features/eu-inference.md Official source: https://docs.lovable.dev/features/privacy-and-security-settings.md ### Data storage: region not confirmed Availability still unclear Business and Enterprise can set a default data hosting region for new projects. The reviewed page does not name which regions are available. According to the Enterprise page, project code lives by default in Lovable's managed infrastructure. Region covered: EU/EEA Documented reason: The sources name no specific region. Official source: https://docs.lovable.dev/features/privacy-and-security-settings.md Official source: https://docs.lovable.dev/introduction/lovable-for-enterprise.md ### Data processing agreement and training With requirements Business and Enterprise include a Data Processing Agreement, and Enterprise can request a custom one. Workspace data on Business and Enterprise is excluded from model training by default, according to the page. On Free and Pro, content may be used for training from 9 September 2026 unless you opt out in your account settings. Region covered: EU/EEA, United Kingdom and Switzerland Documented reason: The sources state no country release for these regions; they describe contract and training rules by plan. Official source: https://docs.lovable.dev/introduction/lovable-for-enterprise.md Official source: https://docs.lovable.dev/features/business/data-opt-out.md ### Availability in Europe Availability still unclear The account page names no country restriction for sign-up. It makes no explicit statement about availability in individual European countries. Region covered: EU/EEA, United Kingdom and Switzerland Documented reason: A missing statement is not availability. Official source: https://docs.lovable.dev/introduction/create-an-account.md ## Accounts & permissions Based on provider documentation. Permissions depend on the account, integration and organization policies. Review the permissions actually requested before use. Configuration covered: Hosted account with workspaces Sources reviewed: 2026-11-05 ### Account used Documented: Sign-up works with email and password, Google, GitHub or Apple. Each account gets a personal workspace on the Free plan. SSO via SAML or OIDC is available only for organisations on Business and Enterprise. Evidence: https://docs.lovable.dev/introduction/create-an-account.md ### Admin approval Documented with conditions: Workspace admins and owners manage privacy and security settings. Some controls, such as EU inference, mandatory two-factor authentication and sensitive data scanning, are Enterprise only. An owner or admin authorises the workspace's Git connection. Evidence: https://docs.lovable.dev/features/privacy-and-security-settings.md Evidence: https://docs.lovable.dev/integrations/git-sync-overview ### Data access & team sharing Documented: Lovable in Slack can only see and change projects the requesting person has access to. Connectors run either on one shared account, personally for you, or with each end user's own account in the published app. Evidence: https://docs.lovable.dev/integrations/lovable-for-slack.md Evidence: https://docs.lovable.dev/integrations/introduction ## Prices and included usage Pricing checked: 2026-11-05 Free plan; Pro from $25 and Business from $50 a month for 100 credits, Enterprise by arrangement. All plans use credits. ### Free 0 US dollars Daily chat allowance, 5 daily build credits (capped at 30 per calendar month), a monthly grant of 20 Cloud and 4 AI credits; Git sync and workspace-private projects; no code editing or download. Price source: https://docs.lovable.dev/introduction/subscription-plans.md ### Pro from $25 a month for 100 credits ($250 billed annually) Tiers from 100 to 10,000 credits; rollover, top-ups, code editing and download, custom domains, user roles and permissions, email support. Price source: https://docs.lovable.dev/introduction/subscription-plans.md ### Business from $50 a month for 100 credits ($500 billed annually) Everything in Pro plus SSO, security center, Lovable API, commit attribution, a Data Processing Agreement and priority support. Price source: https://docs.lovable.dev/introduction/subscription-plans.md ### Enterprise Contact sales Everything in Business plus SCIM, audit logs, organizations, EU inference and dedicated support; without the daily free credits. Price source: https://docs.lovable.dev/introduction/subscription-plans.md According to the page, verified students and teachers get 50% off Pro with 100 credits, billed monthly, for up to 12 months and only for workspaces on the Free plan. The plans page does not make clear whether prices apply per person or per workspace. Pricing details: https://docs.lovable.dev/introduction/subscription-plans.md ## Operation & oversight Based on public provider documentation. Assessments are editorial inferences from the stated properties. Actual time requirements and reliability have not been measured. Deployment covered: Hosted operation in the browser Sources reviewed: 2026-11-05 ### Files and working state Documented: Every edit becomes a version in the project's history, and you can revert to any earlier version. Deleted projects can be restored for 60 days. Memory is stored separately from project files. Evidence: https://docs.lovable.dev/features/projects/overview Evidence: https://docs.lovable.dev/features/memory.md ### Sign-ins Documented with conditions: You sign in with email, Google, GitHub or Apple; SSO is available from Business. On Enterprise a workspace can require two-factor authentication for everyone. How long sign-ins stay valid was not established in this review. Evidence: https://docs.lovable.dev/introduction/create-an-account.md Evidence: https://docs.lovable.dev/features/privacy-and-security-settings.md ### Starting further runs Documented: In Build mode, Lovable can work on one message for up to 10 hours. Jobs run scheduled background tasks in the backend, for example every day at 08:00; Slack messages also start work. The Jobs page does not say which plans include Jobs. Evidence: https://docs.lovable.dev/features/agent-mode Evidence: https://docs.lovable.dev/features/jobs Evidence: https://docs.lovable.dev/integrations/lovable-for-slack.md ### Human intervention Documented: In Build mode Lovable applies changes directly; the page describes no approval of changes there, only credit check-ins that pause long messages. In Plan mode you edit and approve a plan before building. The stop button halts the current task. Evidence: https://docs.lovable.dev/features/agent-mode ### Detecting errors Documented with conditions: Each job run appears with the status Succeeded, Failed or Running; the view does not update live. Project monitoring can alert you when jobs fail. The Jobs page describes no automatic retries. When credits run out, building stops. Evidence: https://docs.lovable.dev/features/jobs Evidence: https://docs.lovable.dev/introduction/credits-and-usage.md ### Resuming after failures Documented: Undo latest edit discards changes after a stop; completed changes otherwise remain. You restore earlier versions from the history and deleted projects for 60 days from Recently deleted. Evidence: https://docs.lovable.dev/features/agent-mode Evidence: https://docs.lovable.dev/features/projects/overview ### Editorial assessment Lovable works in the browser with version history and undo; recurring work runs through backend jobs that consume credits. Plan credits, roles and checking the published app yourself. ## Memory and conversation continuity Scope: Project and personal memory, instructions and AGENTS.md Sources accessed: 2026-11-05 Lovable saves notes automatically and lets you view them in the project and workspace settings and change them through chat. ### Storage and scope Documented: Project memory is stored separately from your project files. Technical and architecture decisions go into AGENTS.md files in the project. Personal memory belongs to your workspace account. Instructions are separate text fields at project and workspace level. - [Lovable: Memory](https://docs.lovable.dev/features/memory.md) - [Lovable: Knowledge](https://docs.lovable.dev/features/knowledge.md) ### Selection and timing Documented: Lovable saves preferences, decisions and facts you state automatically while you build. You can also ask in chat to remember or forget something. Memory is on by default on all plans; on paid plans Dreaming tidies it in the background. - [Lovable: Memory](https://docs.lovable.dev/features/memory.md) ### Inspect, edit and delete Documented: You see project memory under More, Settings, Memory, and personal memory under workspace settings. Both show a summary Lovable writes. You change passages with Change in chat, and to delete something you ask Lovable in chat to forget it. - [Lovable: Memory](https://docs.lovable.dev/features/memory.md) ### Agent and channel boundaries Documented: Everyone who can edit the project shares project memory; others cannot read personal memory. It can be switched off per workspace and set per project to you and project, just project or nothing. Chats outside projects are private, including from admins. - [Lovable: Memory](https://docs.lovable.dev/features/memory.md) - [Lovable: Chats](https://docs.lovable.dev/features/chats.md) ### Retrieval Partly documented: Instructions are always in context, while skills are loaded on demand. Lovable reads AGENTS.md and CLAUDE.md at the top level with every message. For chats outside projects, the page says Lovable recalls earlier preferences only for you and not in every case; better to reference a chat with @. The pages do not describe how memory is retrieved otherwise. - [Lovable: Knowledge](https://docs.lovable.dev/features/knowledge.md) - [Lovable: Chats](https://docs.lovable.dev/features/chats.md) ### Session reset and compaction Partly documented: In very long chats outside projects, Lovable can lose access to the oldest content; the page suggests keeping key material in workspace instructions or a file. Dreaming does not run in projects with Git sync, in workspaces with EU inference, or in projects set to nothing. The reviewed pages say nothing about compaction in the project chat. - [Lovable: Chats](https://docs.lovable.dev/features/chats.md) - [Lovable: Memory](https://docs.lovable.dev/features/memory.md) ## Context An account and credits. On the Free plan you cannot edit or download code. You are responsible for your app's security. Provider documentation, pricing, privacy and trust pages reviewed. Example tasks are editorial; hands-on tests are still pending. ## Open points Europe, availability still unclear: - Data storage: region not confirmed - Availability in Europe ## Sources - https://docs.lovable.dev - https://docs.lovable.dev/introduction/subscription-plans.md - https://docs.lovable.dev/introduction/credits-and-usage.md - https://docs.lovable.dev/features/eu-inference.md - https://docs.lovable.dev/introduction/getting-started.md - https://docs.lovable.dev/features/privacy-and-security-settings.md - https://docs.lovable.dev/features/business/data-opt-out.md - https://docs.lovable.dev/features/chats.md - https://docs.lovable.dev/features/projects/overview - https://docs.lovable.dev/features/projects/settings.md - https://docs.lovable.dev/integrations/git-sync-overview - https://docs.lovable.dev/features/jobs - https://docs.lovable.dev/features/agent-mode - https://docs.lovable.dev/glossary - https://docs.lovable.dev/features/knowledge.md - https://docs.lovable.dev/features/memory.md - https://docs.lovable.dev/introduction/lovable-for-enterprise.md - https://docs.lovable.dev/introduction/create-an-account.md - https://docs.lovable.dev/features/security - https://lovable.dev/terms - https://docs.lovable.dev/integrations/lovable-for-slack.md - https://docs.lovable.dev/integrations/introduction ## Provider According to its terms of service, Lovable Labs Incorporated provides Lovable. The terms are governed by Delaware law. - Operator: Lovable Labs Incorporated - Organization: Provider - Website: https://lovable.dev - Reviewed: 2026-11-05 - Provider source: https://lovable.dev/terms - Provider source: https://docs.lovable.dev [Sources and method](https://ki-agentenkompass.de/en/sources/)