STOCKHOLM—In the frenetic world of artificial intelligence, few startups have scaled as explosively as Lovable.dev. Founded in late 2023 by Anton Osika and Fabian Hedin, the Swedish outfit lets users describe apps in plain English, spitting out full-stack software powered by models from OpenAI and Anthropic. What began as a tool for non-coders has ballooned into a platform claiming 8 million users, $200 million in annual recurring revenue, and over 100,000 daily projects by late 2025.

Investors can’t get enough. A $330 million Series B in December 2025, led by Alphabet’s CapitalG and Menlo Ventures with Nvidia’s NVentures joining, pegged Lovable at $6.6 billion—more than tripling its $1.8 billion valuation from a mere five months prior, according to Reuters. Earlier, Accel anchored a $200 million round, calling it Europe’s largest for Stockholm. CEO Osika, speaking at Web Summit, touted the platform as “the last piece of software,” handling everything from user insights to deployments.

From GPT Engineer to Global Phenomenon

Osika’s roots trace to GPT Engineer, his 2023 open-source project that used large language models for app coding. Lovable built on that, bundling React, Tailwind, and Supabase with one-click deploys and built-in cloud infra. Non-technical founders now ship ideas “in days, not months,” the company pitches on its site. Users include Klarna, Uber, and Zendesk, slashing prototype times from weeks to hours—one healthcare firm visualized patient journeys for invoices after a nurse’s suggestion.

By March 2025, Contrary Research pegged Lovable at 10.4 million monthly visits, outpacing Replit and Bolt.new, with 45,000 paid customers and $50 million ARR by May. Forbes dubbed it “the fastest-growing software startup ever,” hitting $100 million ARR in eight months. Enterprises like Deutsche Telekom weave it into workflows for faster alignment.

Laela Sturdy of CapitalG gushed: “Lovable has done something rare: built a product that enterprises and founders both love.” Yet beneath the hype, insiders whisper of cracks in the foundation.

Rocket Fuel for Prototypes, Perils for Production

For ideation and MVPs, Lovable shines. “Speed is king,” notes a Medium analysis by AOI::The Art of AI, praising conversational AI for rapid starts without teams. Entrepreneurs like Theresa Anoje built job sites sans coding expertise; Airweave hit $17,000 monthly revenue post-Lovable prototype, securing $6 million funding, per Forbes.

TechCrunch reported nearing 8 million users by November 2025, up from 2.3 million in July, with Osika eyeing corporate expansion. The platform’s Agent mode mimics senior devs, breaking tasks, reviewing code, and iterating—ideal for solo builders or small teams.

But scale exposes limits. Users complain of iteration woes: “Every broken refactor and wasted credit chips away at trust,” tweeted Kuruzo. Reddit threads echo frustrations—AI loops on bugs, burns credits on retries, ignores prompts for complex logic. Superblocks’ 2026 review: excellent for prototypes, but “security and data handling feel immature, debugging traps you in prompting loops.”

Terms of Service: Ownership Wins, Liabilities Lurk

Lovable’s terms grant users ownership of apps and AI outputs, subject to third-party model rights—a boon for startups commercializing MVPs, per the official Terms of Service. A limited license lets teams build without IP surrender, and GitHub sync eases exports.

Yet risks abound. “AI Output may contain errors… do not rely on it for critical functions,” the terms warn. Liability caps at 12 months’ fees; no coverage for indirect damages like downtime from third-party providers. Users indemnify Lovable for misuse, and non-PII data fuels their training—exposing proprietary prompts. Subdomains like lovable.app can be reclaimed sans refund.

A Medium deep-dive flags stage-specific pitfalls: ideation speed versus growth dependencies. Reddit user Kind-Arachnid3443 decried autonomous backend migrations to Lovable Cloud without consent, dubbing it unfit for production sans backups.

Security Scares and Vendor Ties

Early 2025 brought headaches. Replit’s Matt Palmer exposed Row Level Security flaws in Lovable apps, enabling data leaks via misconfigured Supabase. Many sites left databases public. Lovable responded with scanners detecting RLS misuse, code reviews, and API key warnings, per CEO Osika’s X post hiring security lead Igor.

Trustpilot gripes include random app deletions, ignored support, unrequested edits. “All sites injected with Lovable tagger, spyware… 70% malicious,” claimed Reddit’s int-gambler. Vendor lock-in irks: migrating from Lovable Cloud to external Supabase proves tricky, per X users. Outages hit hard—January 2025 downtime prompted credit refunds and free weekends.

Osika insists: “We’re making Lovable the most secure place to build.” SOC 2 and ISO certifications followed, blocking 1,000 risky projects daily. Still, Barclays noted 40% traffic dips by September 2025 amid sustainability doubts.

Rivals Circle: Replit, Cursor, Bolt.new Duel

Lovable leads vibe coding but faces heat. Cursor, pro-dev focused, hit $30 billion valuation on $1 billion revenue. Replit offers 30+ languages, robust IDE; Bolt.new flexes frameworks like React Native. v0 excels React UIs with Vercel ties. A Medium comparison: Lovable tops full-stack for beginners, but Replit wins professionalism, v0 zero lock-in.

“Replit is a country mile better… less black box,” per Reddit. Windsurf suits teams; Lovable shines client demos but falters scalability. Sacra warns of positioning tension serving coders and noobs alike, risking mediocrity versus specialists.

Forbes notes rivals like StackBlitz and Cognition nip heels, plus Big Tech—OpenAI, Anthropic launch tools using their own models Lovable rents.

Enterprise Push Amid Execution Hurdles

Lovable eyes corporates, adding collaboration and security hires. Uber cut design testing to five days; Lumoo, an AI fashion app, reached $800,000 ARR in nine months. Yet G2 reviews praise scalability for budgets but flag complexity breaks.

Challenges persist: credit models drain on refactors ($29/month starter, tokens-based), inconsistent support. X threads lament modals reloading on tab switches, auth bugs. Agencies laud MVPs but need devs for polish, per MCStarters.

Osika: market “big enough for multiple winners.” With $6.6 billion bet, Lovable must prove vibe coding scales beyond hype—or cede ground to entrenched foes.

Path Forward: Balancing Hype with Hard Reality