SAN FRANCISCO — Two years after quietly launching a privacy-focused AI platform without outside investors, Venice AI has raised $65 million in its first external round at a $1 billion valuation, becoming a unicorn before most of Silicon Valley had heard its name.
The Series A, led by Dragonfly, the crypto-focused venture firm, and joined by Coinbase Ventures and North Island Ventures, was announced Wednesday. Venice reached unicorn status on its first fundraise, unusual in any year, and particularly striking given that most AI companies in 2026 spent years in multiple funding rounds before their valuations caught up to their revenues, TechCrunch reported.
Venice’s revenue caught up first. The company says it is already profitable, with annualized run-rate revenues exceeding $70 million from subscriptions and API access across three million active users who generate 1.7 million API calls daily. None of that happened with institutional money, which meant Venice spent its first two years answering to users rather than investors.
The platform gives users access to more than 200 AI models, including open-source alternatives the company describes as “uncensored.” All user input is encrypted client-side and routed through external proxies, with Venice retaining nothing on its own systems. The effect, in practice, is an AI assistant that has no record of who asked what, not by policy but by architecture.
That distinction has become commercially significant. The mainstream AI industry has spent the same period building consumer products that, by design, collect conversation data to improve their models. OpenAI, Google, and Meta retain user interactions by default. Venice is a bet that enough people, particularly in markets where digital privacy carries legal force, will pay for a fundamentally different arrangement.
The funding’s origin says as much as its amount. Dragonfly built its reputation backing blockchain infrastructure. Coinbase Ventures invests strategically in anything extending the crypto economy’s reach. Neither typically leads Series A rounds in AI consumer platforms. That they did here reflects something specific about Venice’s founding logic: CEO Erik Voorhees built his career making financial systems more decentralized. Voorhees founded Satoshi Dice in 2012 and ShapeShift, the non-custodial crypto exchange, shortly afterward. Venice applies the same distrust of intermediaries to artificial intelligence that ShapeShift applied to money transfer.
The crypto-AI convergence has been accelerating. Eastern Herald reported Wednesday that Robinhood launched its own public blockchain alongside an AI trading agent product, turning its retail brokerage into something that operates more like decentralized infrastructure. The difference is directional: Robinhood is moving an existing consumer base into crypto. Venice is moving a privacy-conscious AI user base toward a crypto-funded product, while barely 8 percent of its users pay in cryptocurrency at all.
“I think it’s actually quite dangerous from a safety perspective, for the world to enter this next phase and have everyone be constantly watched,” Voorhees said in a statement to TechCrunch. His company’s answer is architectural: making surveillance of its users technically impossible rather than contractually prohibited. That distinction is what Venice is selling.
There are regulatory variables Venice has not yet confronted at scale. The platform’s “uncensored” model framing, which allows users to access outputs that mainstream AI platforms restrict, is likely to come under scrutiny in Europe and Australia, where content moderation obligations on AI providers are moving through legislation. The EU AI Act and Australia’s proposed AI framework both contemplate obligations on operators of high-risk AI systems, and Venice’s claim that it cannot see what users generate does not necessarily exempt it from those obligations. The company has not disclosed how it plans to handle compliance in those markets as it scales.
The $65 million partly funds GPU infrastructure to reduce Venice’s dependence on leased compute. The company has also layered a crypto-token economy on top of its product: the VVV token launched in January 2026, with users able to stake it to generate DIEM, redeemable for daily AI credits. Only 8 percent of users currently pay in crypto, suggesting the token system is an experimental revenue channel rather than a core product feature.
Meta launched its own incognito AI chat mode in May, a concession that users were paying attention to where their conversations went. The feature is opt-in, and Meta’s data practices for users who do not enable it remain unchanged. Venice’s position is that optional privacy is not really privacy. The $70 million in annual revenue says enough users agree to build a business around that argument. What Venice does not yet know is whether they stay when the incumbents offer better opt-outs, or when the regulators it has so far avoided finally arrive.
