$1 billion.

That's the valuation of Venice AI -- the "privacy-first, uncensored" AI platform -- in its latest funding round. The $65 million Series A was led by crypto VC Dragonfly with participation from Coinbase Ventures and others -- more importantly, this barely two-year-old company is already profitable, with annualized revenue exceeding $70 million.

Privacy Isn't a Niche Need -- It's a Mass-Market Pain Point

Venice AI's numbers speak volumes: 850,000 unique website visitors, 3 million monthly active users, 1.7 million daily API calls. Users can access over 200 AI models on the platform (including open-source models and routed access to closed models like OpenAI and Anthropic). All user inputs are encrypted client-side and routed through external proxies before processing; Venice's own systems store zero user data. Some models offer end-to-end encryption, though that requires a paid subscription.

Founder Erik Voorhees's background is fascinating -- an early Bitcoin evangelist who founded the Bitcoin gambling site Satoshi Dice and crypto exchange ShapeShift, and a long-time advocate for user privacy. He positions Venice as a "neutral tool and neutral platform," consistent with Bitcoin's philosophy: "I think a world moving into its next phase where everyone is under constant surveillance is actually far more dangerous from a security perspective -- far more dangerous than someone asking a controversial question."

Venice's growth trajectory is telling. The founder admits that at launch the product's capabilities were far behind ChatGPT, and users used it purely for privacy; but now the functionality is very close to ChatGPT, and once the experience gap closes, "not being watched" transforms from a niche geek demand into a mass-market differentiator.

Models on the platform cover text, image, audio, and video generation. Users can freely choose models with different performance levels, quality, and censorship thresholds, and can even customize AI personas. Venice also modifies open-source models' system prompts to make them respond more openly, without adding extra restrictions.

The Third Path to AI Commercialization

Venice AI's success points to a third path for AI commercialization.

Over the past year, the most discussed AI monetization paths have been either consumer subscriptions (ChatGPT Plus, Claude Pro) or enterprise APIs/services. Venice proves that "value-proposition-driven vertical platforms" can work just as well -- you don't need to build a smarter model than GPT; you just need to go all-in on one value dimension (in this case, privacy and freedom) and get the experience close enough to mainstream products, and users will pay.

Notably, despite having crypto roots (it issued two tokens, VVV and DIEM; users can stake VVV to mint DIEM for AI credits), Voorhees revealed that only about 8% of users pay with cryptocurrency. This means Venice's core user base isn't crypto natives -- it's a broader audience of ordinary users who care about privacy.

Post-funding, Venice plans to use the capital to buy GPUs, build its own data centers, and stop leasing GPUs to improve gross margins. This follows the same logic as vibe coding platforms like Base44 building their own models -- when user scale and call volume reach a certain level, the ROI of owning your infrastructure turns positive.

Of course, "uncensored" is a double-edged sword. AI mental health risks, deepfakes, and misinformation are real problems, and completely unrestricted platforms can easily be abused. Venice's response, like many crypto projects: tools are neutral, responsibility lies with the user. That position is ethically controversial, but from a business standpoint, it clearly taps into a real and sizeable market.

OpenAI's ChatGPT adoption data, released the same day, shows ChatGPT usage continuing to grow globally across more languages and regions. But Venice's $1B valuation and $70M ARR demonstrate: not all users want a "politically correct" AI monitored by big corporations that might refuse to answer because you asked a sensitive question. The AI market isn't winner-takes-all -- it stratifies, just as the internet has both Google and Tor, both Facebook and Signal.