Venice AI just raised $65 million and turned into a billion dollar company on a pitch that runs against almost everything Silicon Valley's biggest AI labs are built on: don't log what the user typed.
Erik Voorhees spent the last decade building crypto companies on the premise that money doesn't need a middleman watching every transaction. First it was Satoshi Dice, the bitcoin gambling site. Then ShapeShift, the exchange that let people swap crypto without creating an account. Now, with Venice AI, he's applying the same logic to chatbots. According to TechCrunch, the startup closed a $65 million Series A led by crypto venture firm Dragonfly, with Coinbase Ventures and North Island Ventures also participating, pushing its valuation to $1 billion. It's Venice's first outside fundraise since Voorhees founded it two years ago.
The number that ought to make other AI founders pay attention isn't the valuation. It's that Venice is already profitable. Voorhees told TechCrunch the company is running at an annualized revenue rate north of $70 million, without having taken a dollar of venture money until now. That's rare in an industry where compute costs routinely bury revenue before it reaches the bottom line.
Venice's product is straightforward to describe and harder to build. The platform gives users access to more than 200 AI models, mixing open source systems that Venice hosts on its own data centers with routed queries to closed models from OpenAI and Anthropic. Input gets encrypted client side, passed through an external proxy, processed, and returned, with nothing stored on Venice's own systems afterward. TechCrunch reports the site now pulls more than 850,000 unique visitors and serves over 3 million active users, generating roughly 1.7 million API calls a day.
You don't get that kind of traction by being merely private. Venice also markets itself as uncensored, meaning it won't refuse a prompt the way ChatGPT or Claude might over safety guidelines. That's the more contested half of the pitch. Big labs build in content restrictions deliberately, often after painful public incidents. Venice is betting that a meaningful slice of users, businesses, and developers building autonomous agents want those restrictions removed and are willing to pay for a service that doesn't retain their prompts in order to make that call later.
The investor list tells you who actually believes this. Dragonfly and Coinbase Ventures aren't AI native funds chasing the next foundation model. They're crypto investors who watched Voorhees launch a token called VVV on Coinbase's Base network back in January. VVV has a hard cap of 100 million tokens, half of which went to roughly 100,000 Venice users and AI community projects through an airdrop. Stake VVV and you can mint a second token, DIEM, which generates a dollar's worth of Venice AI credits every day you hold it. It's a loyalty program wrapped in token economics, and it's also how Venice built a user base without buying Super Bowl ads.
Voorhees says the fresh capital is going toward hardware, not headcount. Venice currently leases its GPUs. The plan now is to buy them outright and start building its own data centers, which would let the company escape the margin squeeze that comes with renting compute from cloud providers at a markup. That's not a glamorous use of $65 million, but it's the one move that actually protects the profitability Voorhees is bragging about.
Zoom out and the timing isn't an accident. Regulators are turning up scrutiny on how AI companies handle user data, and that pressure hasn't gone unnoticed among founders looking for an angle that doesn't require out competing OpenAI on raw model quality. Frankly, the industry default of training on everything you type, and storing it indefinitely, was always going to invite a competitor selling the opposite. Venice's bet is that trust becomes the thing customers actually pay for once every lab has access to roughly the same underlying models anyway.
Whether a $1 billion valuation holds depends on whether that trust argument survives contact with regulators who might not love an uncensored AI platform as much as crypto VCs do. For now, Venice has something most AI startups its size don't: real revenue, a profitable business, and a story that doesn't require anyone to believe next year's model will finally justify the spend.
Also read: Cisco is betting its whole workforce on AI agents while it cuts jobs • Meta Is Quietly Building a Business to Sell Its Leftover AI Computing Power • Twelve Labs raises $100 million as Amazon bets its Trainium chips on video AI
