Venice AI raised a $65 million Series A at a $1 billion valuation, its first outside round, on the back of $70 million in annualized run-rate revenue and 3 million active users. Crypto-focused Dragonfly led the deal, with Coinbase Ventures and North Island Ventures participating. The company, founded by Erik Voorhees, sells a privacy-first alternative to mainstream chatbots and is already profitable.
The traffic numbers explain the check size. Venice draws more than 850,000 unique visitors to its website and processes an average of 1.7 million API calls per day. It offers access to more than 200 AI models — hosting open-source ones on its own infrastructure and routing queries to closed models from OpenAI and Anthropic through an external proxy. No user data is stored on Venice's systems, and paid tiers add end-to-end encryption on select models.
The pitch is straightforward: use frontier AI without a faceless platform logging every query. Input is encrypted and decrypted client-side, and the routing layer strips identifiers before requests hit a model. That architecture is expensive to run but is now the product moat as competitors invest in content filters, identity checks, and safety guardrails that a chunk of the market clearly does not want.
Key facts
- 01Venice AI raised a $65M Series A at a $1B valuation, its first external round, led by Dragonfly with Coinbase Ventures and North Island Ventures participating.
- 02The company is profitable at over $70M in annualized run-rate revenue, with 3M active users and 1.7M API calls per day.
- 03Venice offers access to more than 200 AI models, routing queries through an external proxy with no data stored on its own systems.
- 04Only about 8% of users pay in crypto, despite Venice launching the VVV token in January and DIEM the previous August.
- 05New capital will fund GPU purchases and Venice's own data centers to end GPU leasing and lift gross margins.
Voorhees is an early Bitcoin advocate whose prior companies include the gambling site Satoshi Dice and the exchange ShapeShift. He frames Venice as a neutral platform in the same tradition, and pushes back on the argument that broad access is inherently unsafe.
“This is the same principle that you have in Bitcoin, where Bitcoin, as a neutral protocol, works the same way for all people.”— Erik Voorhees, Venice AI CEO
The safety debate matters because Venice explicitly markets an "uncensored" experience. Users pick from models generating text, images, audio, and video, with performance and filtering varying by choice. The site foregrounds AI "characters" that users can customize and chat with, and Voorhees said Venice tunes some open models' system prompts to answer more openly without adding new restrictions on top.
Crypto is woven into the business but is not the business. Venice launched the VVV token in early January and added a second token, DIEM, the previous August. Users buy VVV, stake it to mint DIEM, and receive $1 worth of AI credits per day to spend on the platform. Only about 8% of users actually pay with crypto — the rest transact in normal currency, and Voorhees credited the bigger growth driver to Venice closing the capability gap with ChatGPT.
That gap has narrowed sharply. Two years ago Venice was a distant substitute that people tolerated for the privacy story. Today the underlying open models, plus routing to closed frontier models, put it within striking distance of the default chatbot experience.
The fresh cash goes toward vertical integration on the compute side. Venice currently leases GPUs; Voorhees said the plan is to buy GPUs outright and build its own data centers, which should lift gross margins meaningfully at 1.7 million API calls per day and rising. That is the same playbook every mid-scale inference operator is running, but Venice enters it already profitable rather than burning to hit scale.
The risk column is not empty. An uncensored, privacy-preserving AI service will inevitably be linked to output the wider industry has spent two years engineering around, and the regulatory posture toward chatbots is tightening rather than loosening. Venice's defense — that it is a neutral protocol operator, not a publisher — has worked for Bitcoin-adjacent businesses in some jurisdictions and failed in others. A lawsuit, a state-level content bill, or a payments-side cutoff could all reprice the model quickly.
The strategic read is that Venice has found a durable niche the frontier labs cannot serve without cannibalizing their enterprise and government contracts. Every safety commitment OpenAI, Anthropic, and Google make to Fortune 500 buyers and regulators is a customer Venice can convert on privacy and permissiveness grounds. At $70 million ARR and profitable, it is no longer a curiosity — it is a proof that a meaningful slice of AI demand will pay a premium to not be watched, and that slice is now backed by a unicorn valuation and crypto-native capital that is unlikely to push it toward the mainstream compromise.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
