Lovable, one of the darlings of the vibe-coding world and one of Europes fastest-growing AI startups, is working on raising a fresh round of over $150 million at a near $2 billion assessment, the Financial Times reports.The raise and giant action up in appraisal comes just months after the Swedish startup raised a $15 million round led by Creandum in February.
The company described that round to A Technology NewsRoom as pre-Series A, however with numbers this big, its safe to say that Lovable has actually jumped from seed rounds to priced development rounds, whatever the serial alphabetic label should be.
Accel is said to be leading this new raise, with Creandum and others like 20VC participating.While the business is technically two years old, established in 2023, it launched its web app-building product in late November.
In May, Lovable CEO Anton Osika tweeted that Lovable hit $50 million in ARR in six months.Lovable, like competitors Replit and Bolt, builds whole web apps from an initial text timely, consisting of a user interface/front end (frequently via the popular UX coding tool React) and linked to a database like Supabase.
Some users say its cost effective, beginning at $25 a month for 250 credits.
One Reddit user documented an app with 29,000+ lines of code and dozens of functions developed for $250.
On Monday, Lovable revealed that it was releasing a beta variation of an AI representative that might automate more jobs like editing code after checking out job files or debugging.
Adorable will charge on a usage-based model for this: The more the representative is asked to do, the more credits it will charge.While this may increase costs for users if they turn over their app management to the agent, this rates model is forming up to be the default organization model for representatives.
This is because the AI start-ups themselves have to pay variable charges to model providers like OpenAI or Anthropic.
All this to say, such business model methods would make investors happy.Accel, 20VC, and Lovable did not react to a request for remark.