Lovable's run rate passes $600M as vibe coding goes enterprise
Lovable's revenue line is now an enterprise story, a White House memo turns AI safety into a loyalty test, and a startup bets $20 million that agents belong in the group chat.
Lovable says its annual run rate has passed $600 million, up from roughly $500 million in June, and the growth is coming from companies rather than hobbyists. Co-founder Fabian Hedin gave the figure on stage at the HumanX summit in Amsterdam this week, saying two-thirds of Fortune 500 companies now use the platform. Named customers include Microsoft, NVIDIA and Deutsche Telekom, which has reportedly built more than 2,000 apps on it. Hedin also said the apps Lovable users have built draw close to a billion visits a month — an order of magnitude more traffic than Lovable's own site. His framing of the product is the interesting part: "Lovable does not output code. The output is a product, and increasingly so, a business."
The number was not a surprise. By August the company was already telling reporters it was tracking toward roughly $600 million by the end of that month, so what Hedin confirmed is a projection landing on schedule rather than a beat. What has changed is the pitch and the price. Lovable has raised more than $700 million across two rounds eight months apart — $300 million at a $6.6 billion valuation last December, then $400 million at $13.3 billion in August, led both times by Menlo Ventures with CapitalG and then the Scaleup Europe Fund. At a $600 million run rate, $13.3 billion is roughly 22 times revenue. That multiple does not price a coding toy; it prices enterprise seats that keep arriving, which is exactly why the Fortune 500 line mattered more than the revenue line.
A memo circulating inside the White House reportedly paints effective altruism as a fringe, cultish movement that "built the AI-doom pipeline," and names Anthropic chief executive Dario Amodei at the movement's foundation. Axios, which obtained the document, says it was written by a Trump political adviser and places Amodei and his sister — Anthropic president Daniela Amodei — inside the network, along with family ties to AI-safety philanthropy. An unnamed administration official told Axios that Amodei "is the embodiment of an ideology and globalist approach to innovation that's counter to the president's America First agenda." Other outlets have picked the story up, but they are reporting on Axios's reporting; the memo itself has not been published anywhere we could verify, so treat the quotes as Axios's account of a document we cannot read.
The timing is the substance. Anthropic is preparing what Axios describes as a record-setting IPO, and the attack lands months before US midterms — which makes this less a debate about AI risk than an attempt to rebrand safety work as un-American ahead of both. It also fits a pattern we have tracked before — Silicon Valley has started punishing the AI doomers — now escalated from investor sniping to a White House document. Amodei has not responded publicly.
Ando came out of stealth with $20 million to build a team messaging platform where AI agents get their own identities instead of being apps you install. Founder Sara Du told TechCrunch the product gives agents inboxes, lets them browse and join channels without being tagged, and includes live calls that can be transcribed and read by an agent. The pitch targets what she calls "meat proxies" — the humans currently relaying agent work between tools and teammates. Accel led the pre-seed, with Index Ventures and Emergence leading the seed; the three investors confirmed the round in their own posts. Ando says it is working with customers in software, real estate and finance across 15 countries, and currently works best for teams of up to about 40 people, with a waitlist beyond that.
The honest read: this is an early-stage entrant in a category the incumbents have already entered. Slack's native bot is an AI agent now, Microsoft has wired Copilot through Teams and Office, and Jack Dorsey's Buzz went after the same idea in July. Du's own admission is the most useful data point — early testers found it "just a jankier messaging platform" until the agent behaviour showed up. Whether agents get first-class identity and permissions inside a company's chat is a real question; whether a 40-person product answers it before Slack does is not settled.
What to watch: whether Anthropic's IPO filing forces the White House memo into the open, and how long Lovable's 22x multiple survives if enterprise growth slows.
Is a $13 billion valuation on a self-reported run rate a bet on enterprise seats, or on the vibes? Tell us in the comments.
