In a crypto market full of AI narratives but short on revenue, Venice stands out as one of the few projects able to point to real top-line numbers.
VVV hit an all-time high of $29.19 on Sept. 9, rising more than 40% in 24 hours and pushing its market capitalization above $1.2 billion. The immediate catalyst was NEAR Protocol’s announcement of an integration with Venice around what it described as a private AI and DeFi loop. But the article argues that the deeper force behind the move is the revenue-buyback-burn structure Venice has spent the past 12 months putting in place.
Erik Voorhees said on Aug. 17 that the platform’s annualized revenue had passed $100 million. That came just six weeks after Venice disclosed a $70 million annualized figure during its July Series A round. The round was led by Dragonfly, raised $65 million, carried a $1 billion valuation, and included participation from Coinbase Ventures.
Those numbers make VVV look like a rare AI-linked token with actual business activity behind it. After the rally, however, the article says the market needs to answer a simpler question: can Venice’s buybacks and burns outrun token emissions?
Three burn channels now feed the model
Venice’s token burn system currently runs through three channels.
The first is a discretionary monthly buyback. Since December 2025, Venice has used part of platform revenue to buy VVV on Aerodrome, a decentralized exchange on Base, and send the tokens to a burn address. The amount has increased month by month since launch. The latest buyback, on Sept. 9, was about $391,000. So far, cumulative monthly buyback burns total about 217,000 VVV.
The second is subscription-triggered burns. Since April 2026, every new paid subscription has automatically triggered VVV buybacks and burns: $2 for Pro, $5 for Pro Plus, and $10 for Max. The article stresses two words here: automatic, and new subscriptions. Renewals do not yet trigger burns, though Venice discussed that as a future trigger during a community call on July 22.
The third is API credit purchase burns. Since July 17, 2026, every $100 spent on API credits, including developer usage, has sent $5 into automatic VVV purchases and burns. That ties token destruction directly to platform activity.
Combined, those three channels have returned about $905,000 in value to token holders in 2026. Added to that are 32.5 million unclaimed tokens burned when the airdrop window closed in March 2025, plus about 1 million tokens repurchased and burned by the team. In total, the on-chain burn address now holds about 33.72 million VVV, equal to roughly 42.3% of the effective circulating supply.
Emissions are falling fast, but dilution still matters
Burns are only one side of the equation. The other is how quickly new tokens enter the market.
VVV’s annual emissions stood at 14 million tokens when the token generation event took place in January 2025. Venice began a systematic emissions reduction in March 2026, cutting the number month by month: to 6 million before June, to 5 million on May 1, to 4 million on June 1, to 3 million on July 1, and then to 2.5 million on Sept. 1. Another cut, to 2 million, is scheduled for Oct. 1.
From 14 million to 2 million, annual emissions have been reduced by about 86%. The article also notes that 100% of emissions go to stakers as yield. There are no team unlocks absorbing emissions, and no foundation allocations dripping into the market.
Even so, price changes the picture. At roughly 2 million to 2.5 million VVV of annual emissions and a token price near $25, the implied yearly sell pressure comes to about $50 million to $62.5 million. For Venice to deliver net deflation, buybacks and burns would need to exceed that amount.
Using $100 million in annualized revenue as the baseline, a 5% burn tied to API credit purchases would imply about $5 million a year in automatic burns. Subscription-triggered burns and discretionary monthly buybacks would add to that. But the article points out that cumulative value returned in 2026 is only about $905,000. Even assuming buybacks rise in step with revenue, the current burn pace still sits well below the dilution pressure created by emissions.
On that basis, the piece argues that net deflation would require either another jump in revenue scale or a much higher buyback ratio. Emissions cuts alone are not enough.
DIEM turns AI compute access into a tradeable perpetual asset
The most imaginative part of the VVV token model, according to the article, is DIEM.
Each DIEM represents $1 of Venice API credits per day, permanently. To mint DIEM, users lock staked VVV, or sVVV. During the lock period, stakers still receive 80% of normal yield, while the remaining 20% goes to Venice. DIEM can be traded on Aerodrome and can also be burned at any time to unlock the original sVVV.
The design shifts AI inference from a pay-per-use model to an ownership model. Hold 10 DIEM, and the holder gets $10 of API credits every day regardless of where VVV trades.
Erik Voorhees compared DIEM to a 30-year mortgage. His point was that once a resource such as AI inference becomes important enough, locking in long-term usage rights at a fixed cost can make economic sense, much like power purchase agreements, airline fuel hedging, or fixed-rate mortgages.
DIEM’s supply cap is being expanded in stages and is expected to rise to 40,000 on Sept. 14. How quickly it grows depends directly on whether VVV holders are willing to lock tokens, and that in turn depends on how they judge Venice’s ability to keep operating over time.
For AI agent developers, the article presents DIEM as a more practical piece of infrastructure. An autonomous on-chain agent could hold DIEM to secure steady inference capacity without repeatedly purchasing API credits or maintaining credit card payment rails. Venice’s documentation already describes the full flow for agent wallets to obtain API keys automatically.
NEAR integration sparked the rally, but on-chain proof still needs to follow
NEAR Protocol’s Sept. 8 integration post on X was the direct trigger for the VVV surge. According to the announcement, users can privately buy VVV through the NEAR network and stake it to gain Venice Pro access, while Venice’s end-to-end encryption runs on NEAR AI Cloud.
In that setup, NEAR fills what the article calls the final piece of Venice’s privacy infrastructure. Venice already says it does not store user data and uses trusted execution environments, or TEE, together with end-to-end encryption to process inference requests. NEAR AI Cloud adds the lower-level encrypted computing layer.
The article is careful on one point: for now, this integration is still more visible in the announcement than in verifiable usage data. Actual VVV trading volume on NEAR, real usage of private purchases, and the scale of Venice inference requests handled by NEAR AI Cloud all still need ongoing on-chain validation.
A single post helping drive a 40% move shows, in the article’s view, how strongly the market is responding to catalysts relative to basic fundamental caution.
The risk list remains substantial
The article says Venice has achieved several things many AI tokens have not: revenue, a working product, users, and on-chain burns that can be verified. Still, it highlights a number of structural risks.
- The VVV contract includes an owner-only mint function. There is no hard on-chain supply cap, and the emissions schedule is entirely set by Venice. VVV is also a purely utility token, with no governance voting rights for holders.
- Token concentration is very high. The top 100 holders control about 98% of supply. Team allocation accounts for 10% with a 24-month vesting schedule, while company operations account for 25% with no public vesting plan.
- The $100 million annualized revenue figure is a run rate, not an audited number. It is extrapolated from a short-term window and does not equal recognized full-year revenue. Venice is under no obligation to disclose revenue composition, and there is no independent audit.
- There is also a yield paradox. While 100% of emissions go to stakers, the drop from 14 million to 2 million tokens means staking APY falls as well. If yields become less attractive, stakers may have to rely on price appreciation to make up the difference, shifting the thesis from income to directional exposure.
The article also notes that VVV has climbed from a low of $0.92 to $29 over the past year, a gain of more than 30x. On that reading, the current valuation already reflects expectations around revenue growth, lower emissions, and the NEAR integration.
What to watch next
The next phase for VVV, the article argues, comes down to a handful of measurable indicators.
First, can monthly buybacks and burns consistently exceed monthly emissions? At current prices, that threshold is about $4.2 million to $5.2 million per month.
Second, can DIEM minting volume and real usage show that AI agents actually want on-chain ownership of compute access?
Third, can the $100 million annualized revenue figure be validated in later quarters as sustained growth rather than a one-time burst?
Venice is trying to answer an old question in crypto: can a token evolve from a speculative instrument into a productive input? DIEM’s perpetual tokenization of AI inference is an elegant design. The article’s conclusion is more restrained. Elegant design still needs equally convincing numbers behind it.
In a market where many AI tokens still offer concept without revenue, Venice at least provides a framework that can be tested. From here, the article says, it becomes a math problem.