Lovable Doubled Its Valuation in Eight Months

Lovable closed a $400 million Series C at a $13.3 billion valuation, announced August 12, 2026. The round was led by Menlo Ventures and the Scaleup Europe Fund — an EU investment vehicle managed by EQT — with participation from existing backers and new investors out of Latin America and Asia, including Tencent.

Eight months ago, in December, the Stockholm company raised $330 million at $6.6 billion. The valuation has exactly doubled. The company was founded in late 2024.

The operating numbers underneath are the reason. Annual recurring revenue has nearly tripled from $200 million and the company says it is tracking toward a $600 million run rate by the end of August. It hosts 60 million projects drawing roughly 900 million monthly visitors — a traffic figure that puts the platform's output, in aggregate, in the same order of magnitude as a top-tier consumer destination.

What Lovable actually sells

You describe software in plain language. It builds it, hosts it, and gives you a URL. That is the whole product, and its apparent simplicity is what makes people underestimate it.

The category acquired the name "vibe coding" as a joke and kept it as a market. The joke assumed the output would be toys — landing pages, hackathon demos, things that break the moment a real user touches them. What actually happened is that the tools got good enough to carry internal tools, client dashboards, marketplace MVPs, and an enormous long tail of software that previously did not exist because no one could justify a developer's month to build it.

That long tail is the business. Lovable is not primarily competing with GitHub Copilot for the attention of professional engineers. It is competing with the decision not to build the thing at all — with the spreadsheet, the manual process, the agency quote that came back at $40,000 and killed the project. Every one of those is a customer who was never in the software market before.

The 900 million monthly visitors number is the proof. Those are not Lovable's users. Those are the users of things Lovable's users made.

The cap table is a geopolitical document

Read the investor list again: an EU sovereign-adjacent fund and Tencent wrote into the same round.

The Scaleup Europe Fund exists because Brussels spent the last three years watching every European technology success get acquired, relocated, or diluted into a US cap table before it reached scale. Its explicit purpose is to keep late-stage European companies European. Putting it at the front of a round for the continent's fastest-growing software company is exactly the mandate working as designed — and it arrives in the same month the EU began enforcing the AI Act's transparency regime, which is the other half of the same strategy. Regulate the imports, capitalize the domestics.

Tencent's participation cuts the other way and is the more interesting signal. Chinese strategic capital has been largely frozen out of frontier US AI by export controls, CFIUS review, and the general temperature of the relationship. European application-layer companies are not subject to those constraints in the same way. Lovable is not a model lab; it is a product built on top of models. That distinction is what makes the check possible, and it is a template other investors will notice.

The structural question nobody has answered

Lovable does not train frontier models. It orchestrates them. Its gross margin is therefore a function of somebody else's pricing decisions — and the entire industry spent 2026 driving inference costs down hard. Claude Sonnet 5 was priced specifically to make long agent runs economical. Chinese open-weight models undercut Western APIs by an order of magnitude. Nvidia is reportedly building a trillion-parameter open model to give enterprises something they can run themselves.

For Lovable, all of that is tailwind: input costs falling while pricing power holds. Tripling ARR while your cost of goods deflates is the most flattering possible configuration for a software business.

The risk is the same fact viewed from the other end. If the model providers decide the application layer is where the margin went, they can move into it — and they have. Anthropic ships Claude Code. OpenAI has been layering task execution directly into ChatGPT. Google embeds Gemini in Figma and everywhere else. Each of those is a well-capitalized competitor with a structural cost advantage and no need to earn a gross margin on the model call.

Lovable's defense is that the model is not the product. Hosting, deployment, project persistence, the accumulated non-technical user base, and the 60 million projects already living on the platform constitute switching costs that a better code generator does not automatically dissolve. That is a real moat. It is also a moat made of habit rather than technology, and habits at eight-month-old companies are not yet habits.

What the number means

$13.3 billion on a $600 million run rate is roughly 22x forward revenue for a company growing triple digits with software margins. In the 2026 tape, that is not obviously expensive. It is a bet that the run rate is still early, that the long tail of non-developers building software is enormous, and that Lovable's brand becomes the default verb for it.

Europe has been waiting a long time for a company where the interesting question is whether the multiple is too low. It finally has one.