Written by: Xiao Bing

VVV reached an all-time high of $29.19 on September 9, with a 24-hour increase of over 40%, and the market cap exceeding $1.2 billion. The direct catalyst was the NEAR Protocol's announcement of a "Privacy AI + DeFi closed-loop" integration with Venice, but the underlying logic supporting this surge is the revenue-repurchase-burn mechanism that Venice has gradually built up over the past 12 months.

In a crypto market filled with "AI narratives" yet lacking revenue support, Venice is one of the few projects that can showcase real revenue data.

Erik Voorhees announced on August 17 that the platform's annual revenue exceeded $100 million, just six weeks after disclosing $70 million during the Series A funding in July. This round of financing, led by Dragonfly with $65 million, gave a valuation of $1 billion, with Coinbase Ventures participating as well.

These numbers make VVV appear to be a rare target in the AI track. However, after the surge, the core question that needs a calm answer is: Can Venice's income repurchase burn truly outpace token emissions?

Repurchase and Burn Mechanism: From "Financial Operations" to "On-chain Automation"

Venice's token burn operates on three lines.

One line is the discretionary monthly repurchase. Starting in December 2025, Venice will use a portion of platform revenue each month to buy VVV at market price from Aerodrome (a DEX on the Base chain) and send it to the burn address. The amount on this line has been increasing monthly since it started, with the most recent (on September 9) being approximately $391,000. As of now, about 217,000 VVV have been cumulatively repurchased and burned monthly.

Another line is subscription-triggered burns.

Starting in April 2026, every new paid subscription will automatically trigger the repurchase and burn of VVV: $2 for Pro level, $5 for Pro Plus level, and $10 for Max level. Note that "automatic" and "new subscriptions" are two key phrases. Renewals do not trigger burns for now; this is a future trigger mentioned by Venice in the community conference call on July 22.

The third line is API quota purchase burns. Starting July 17, 2026, for every $100 of API quota purchased by users (including developer calls), $5 will automatically be used to purchase and burn VVV. This line directly ties platform usage to token burns.

Combined, Venice has given back approximately $905,000 worth to token holders in 2026. Additionally, during the airdrop window closure in March 2025, 32.5 million unclaimed tokens were burned all at once, along with about 1 million tokens repurchased and burned by the team, resulting in a cumulative on-chain burn address of about 33.72 million VVV, equivalent to about 42.3% of the effective circulating supply.

Emission Table: From 14 Million to 2 Million

The burn is only half of the equation; the other half is the inflow speed of new tokens.

The annual emission rate of VVV at the TGE in January 2025 was 14 million tokens. Venice began systematic emission reductions starting in March 2026, decreasing monthly:

By June, it decreased to 6 million, down to 5 million on May 1, down to 4 million on June 1, down to 3 million on July 1. The latest step implemented on September 1 has reduced emissions to 2.5 million, and on October 1 it will further decrease to 2 million.

This represents a reduction of about 86% from 14 million to 2 million. Moreover, 100% of the emissions are allocated to stakers as rewards, with no team unlocking to consume emissions and no foundation grants dripping into the market.

Putting both sides together: based on a current emission of about 2 to 2.5 million per year, and calculating at a VVV price of about $25, the market selling pressure corresponding to the annual emissions is about $50 million to $62.5 million. For Venice to fulfill the commitment of "net deflation," the amount spent on income repurchase and burn needs to exceed this figure.

Based on $100 million annual revenue, a 5% API quota burn implies an automatic burn of about $5 million/year, combined with subscription triggers and discretionary monthly repurchases, but the cumulative feedback for 2026 is only $905,000. Even if we assume that repurchase amounts grow in sync with revenue, the current burn speed still falls far short of the dilution pressure from emissions.

Achieving net deflation requires a further leap in revenue scale or a significant increase in the repurchase ratio; simply reducing emissions won't suffice.

DIEM: Turning AI Computing Power into a Tradable Perpetual Asset

The most imaginative part of the VVV token economic model is DIEM.

Each DIEM represents a daily $1 Venice API quota, valid indefinitely. The way to mint DIEM is to lock up staked VVV (sVVV). During the lock-up period, stakers continue to receive 80% of normal earnings, with the remaining 20% going to Venice. DIEM can be traded on Aerodrome or burned at any time to unlock the original sVVV.

This design transforms AI inference capability from a "pay-per-use" model into a "shareholding" model. If you hold 10 DIEM, you have a $10 API quota available daily, regardless of VVV’s market price fluctuations.

Erik Voorhees likens the logic of DIEM to a "30-year mortgage": when a resource (AI inference computing power) becomes significant enough, locking in its long-term usage rights at a fixed cost becomes economically reasonable, just like power purchase agreements, aviation fuel hedging, and fixed-rate mortgages.

The supply cap of DIEM is being phased in, expected to increase to 40,000 on September 14. The speed of its expansion directly depends on the locking willingness of VVV holders, which in turn is based on their judgment of Venice's ability to continue operations.

For AI Agent developers, DIEM may serve as a more practical infrastructure component. An on-chain autonomous Agent can hold DIEM to obtain a stable supply of inference capabilities without the need to repeatedly purchase API quotas or maintain credit card payment channels. Venice's documentation has already described the complete process for Agent wallets to automatically acquire API keys.

NEAR Integration: The Infrastructure Puzzle for Privacy AI

The integration announcement by NEAR Protocol on September 8 was the direct trigger for the surge of VVV. The integration includes: users can privately purchase VVV through the NEAR network and stake to gain access to Venice Pro, and Venice's end-to-end encryption runs on NEAR AI Cloud.

NEAR's role is to complete the privacy infrastructure for Venice. What Venice has already implemented is: not storing user data, using TEE (Trusted Execution Environment) and end-to-end encryption to handle inference requests. NEAR AI Cloud provides the underlying support for encrypted computing.

However, it needs to be clear that this integration is primarily reflected in the announcement stage. The actual trading volume of VVV on NEAR, the real call data for private purchases, and the scale of NEAR AI Cloud's processing of Venice inference requests all require continuous verification of on-chain data.

A single tweet triggering a 40% increase indicates that the market's thirst for catalysts has surpassed its cautiousness regarding fundamentals.

Investment Risks to Watch Out For

Venice has accomplished many things that other AI tokens have not: it has revenue, products, users, and verifiable on-chain burns. However, the following structural risks cannot be ignored:

The VVV contract contains an owner-only mint function. There is no hard cap on supply on-chain, and the emission schedule is entirely unilaterally determined by Venice. VVV is a purely functional token, and holders have no governance voting rights.

Token concentration is extremely high. The top 100 holders control about 98% of the supply. The team allocation accounts for 10% (24 months vesting), and company operations account for 25% (with no disclosed vesting plan).

Annual revenue is an "operational rate" rather than an audited figure. The $100 million annual revenue is extrapolated from a certain short-term window, not representative of confirmed yearly revenue. Venice is not obligated to disclose the composition of revenue nor is there an independent audit.

The paradox of emission yield: 100% of emissions are allocated to stakers, but the drop from 14 million to 2 million emissions means staked APY declines in sync. When yields are no longer attractive, stakers need to rely on price increases to compensate, shifting investment logic from "income investment" to directional bets.

VVV has risen from a low of $0.92 to $29 over the past year, a more than 30-fold increase. The current valuation has fully priced in revenue growth, emission reductions, and the expectations of the NEAR integration.

What will determine VVV's trajectory moving forward are several traceable hard metrics: Can the monthly repurchase and burn amount continue to exceed monthly emissions (currently about $4.2 million to $5.2 million/month)? Can the minting and actual utilization of DIEM prove the demand for on-chain computing power ownership by AI Agents is real? And can the $100 million annual revenue be verified as sustainable growth in subsequent quarters, rather than a one-time surge?

Venice is attempting to answer an age-old question in the crypto world: Can tokens evolve from speculative tools to production factors? The perpetual tokenization of AI inference computing power through DIEM is a beautiful design, but a beautiful design also requires beautiful numbers to support it.

In a market environment where AI tokens are generally "conceptual but have no revenue," Venice at least provides a calculable framework, leaving the rest as a math problem.