Venice.ai is a privacy-first AI platform with 4 million users, a $1.16 billion market cap, and a two-token system that turns AI inference access into a tradeable asset. Founded by Erik Voorhees and backed by a $65 million Series A at unicorn valuation, Venice has burned 33.9% of its genesis VVV supply and slashed emissions by 86%. The question is whether the product’s real traction can close the gap between current inflationary token supply and a promised deflationary future.
What Is Venice.ai?
Venice.ai launched in May 2024 as a private, uncensored alternative to ChatGPT and Claude. The platform aggregates over 200 models across text, image, video, and audio, routing user requests through a privacy proxy that strips identifying metadata before forwarding to external providers. For its own hosted open-source models, Venice operates under a zero-retention contract: prompts are not stored, and chat history stays in the user’s browser by default.
Erik Voorhees, the founder, built ShapeShift into one of the earliest crypto exchanges before dissolving it into a DAO in 2021. The president and CTO is Jesse Proudman, who previously co-founded Makara (sold to Betterment) and Bluebox (sold to IBM).
How Venice Protects Privacy
Venice organizes its privacy architecture into four tiers, each carrying a badge in the UI so users know exactly who can read their prompts.
- Anonymous: Venice strips your identity. The outside model provider still sees the prompt text.
- Private: Venice-hosted open-weights models run with zero data retention.
- TEE: Models execute inside a sealed hardware enclave that even Venice cannot inspect.
- E2EE: Prompts are encrypted on the user’s device and only decrypted inside a verified enclave.
Every Claude, GPT, and Gemini model on Venice is labelled Anonymous. Those providers still receive the raw text. For true privacy, you must use Private, TEE, or E2EE modes. TEE and E2EE capacity is supplied by two external partners, NEAR AI Cloud and Phala Network, running since March 2026. E2EE disables web search, file uploads, and tool use, so there is a real functionality trade-off.
Venice’s own in-house model, Venice Uncensored 1.2, launched in April 2026. It’s a 24-billion-parameter fine-tune of Mistral 24B, built with Dolphin (dphnAI) and trained on Bittensor subnet 4 (Targon), released under Apache-2.0. It has a 128,000-token context window, native vision, and tool use, and is optimized for uncensored and creative use cases. Venice charges roughly $0.20 per 1M input tokens and $0.90 per 1M output tokens for it.
The Two-Token System: VVV and DIEM
VVV is the capital asset of the Venice ecosystem, an ERC-20 token on Base. DIEM is the second token, representing perpetual daily AI compute access.
Staking VVV earns yield from token emissions and unlocks Venice Pro access at 100 VVV staked. Locking staked VVV lets you mint DIEM. Each DIEM provides $1 of Venice API credit per day, forever. Unused credit does not roll over.
DIEM uses an exponential mint curve. At current supply of roughly 36,696 DIEM against a 40,000 target, minting one new DIEM requires locking approximately 438 sVVV as recoverable collateral. The curve steepens as supply approaches the target, creating natural scarcity. At the current price of roughly $1,612, the implied discount rate is about 23%, meaning the market prices in moderate confidence that Venice will keep delivering that $1/day.
Tokenomics Design
VVV launched on January 27, 2025 with a 100 million token genesis supply and no presale. The allocation breakdown:
| Category | Percentage | Notes |
|---|---|---|
| Airdrop | 50% | 25M to 100,000+ Venice users; 25M to Base AI-community protocols |
| Venice AI | 35% | Company grant; 10M of it to team (25% upfront, rest streaming 24 months) |
| Incentive Fund | 10% | Ecosystem grants |
| Liquidity | 5% | Aerodrome LP on Base |
Source: Venice AI Blog
The airdrop unclaimed tokens were burned in March 2025, the single largest burn event. The team also burned VVV in the same event. Together, these one-time events account for ~99% of all VVV burned to date.
| Metric | Value |
|---|---|
| Circulating Supply | 48.82M VVV |
| Total Supply | 81.15M VVV |
| Max Supply | Uncapped |
Source: CoinGecko, VeniceStats
The max supply is uncapped because staking emissions continue, but Venice has aggressively reduced the emission rate from 14M VVV at launch to 2M VVV as of October 2026, an 86% cut. The schedule stepped down monthly through 2025 and 2026.
Vesting uses on-chain Sablier streams with auditable cliff dates. A critical event arrives in January 2027, when 12 vesting streams end simultaneously, cutting the daily drip rate by roughly 94%.
Market Data
VVV has held up reasonably well in a volatile AI-crypto sector. The token trades at roughly half its all-time high but has recovered over 2,500% from its December 2025 low.
Token Metrics as of October 19, 2026
| Metric | Value |
|---|---|
| Price | $23.95 |
| Market Cap (Circulating) | $1.16B |
| Market Cap (FDV) | $1.94B |
| 24h Volume | $32.18M |
| Holders (Base) | 156,289 |
Performance
| Period | Change |
|---|---|
| 24h | +6.4% |
| 7d | -4.7% |
| 30d | -12.9% |
| 1y | +1,498% |
Source: CoinGecko
The 67% gap between circulating market cap and FDV is a real overhang. It reflects the 33.2M VVV difference between circulating and total supply, most of it in the treasury reserve, incentive fund, and remaining vesting.
Top Exchanges by Volume
| Exchange | 24h Volume | Type |
|---|---|---|
| Aerodrome Slipstream 3 | $4.68M | DEX |
| Bybit | $4.24M | CEX |
| Gate | $3.81M | CEX |
| Coinbase | $3.69M | CEX |
| Kraken | $2.33M | CEX |
Source: CoinGecko Tickers
VVV is primarily a Base-native token with deep DEX liquidity on Aerodrome and major CEX listings on Coinbase, Bybit, Gate, and Kraken.
The Burn Mechanics
Venice runs two burn channels: discretionary buybacks funded by revenue (running since November 2025) and programmatic burns triggered by qualifying events. Every new Pro subscription burns $2 of VVV, Pro+ burns $5, and Max burns $10. Since July 2026, $5 of every $100 in API credit purchases also triggers a burn.
The combined ongoing burn rate is approximately 31,833 VVV per month. That is real, but it is dwarfed by the roughly 571,630 VVV in monthly inflows (166,780 from emissions plus 404,850 from vesting). Venice is net inflationary at current rates.
The thesis depends on two things: the January 2027 vesting cliff that removes 94% of the drip, and revenue growth that scales the burn. Venice claims no outside capital before the Series A, which suggests the subscription and API business was already generating meaningful revenue.
Social Sentiment and Community
- X (Twitter): @AskVenice has approximately 93.9K followers with regular product updates.
- Discord: discord.gg/askvenice has 16,909 members with 1,896 online.
- Telegram: The official channel has only 390 subscribers; community primarily lives on Discord.
- GitHub: The veniceai org has 11 public repositories and 281 followers. The api-docs repo has 67 stars with recent activity.
Community sentiment is mixed-to-positive. Supporters highlight the real user base, aggressive emissions cuts, and Voorhees’ track record. Critics point out that closed models still receive prompt text, the “uncensored” branding has limits, and Venice is ultimately a centralized company running a proxy, not a decentralized network.
Recent Developments
- July 2026: Venice raised a $65M Series A at $1B valuation, led by Dragonfly with Coinbase Ventures, F-Prime, and North Island Ventures participating. This was the company’s first outside capital.
- July 2026: Added programmatic credit burns where $5 of every $100 in API credit purchases buys and burns VVV.
- July 2026: Emissions reduced from 3M to 2M VVV/year, continuing the stepped reduction schedule.
- April 2026: Launched programmatic subscription burns where every new Pro, Pro+, and Max subscription triggers automatic VVV buy-and-burn.
- March 2026: Venice became a recommended model provider for OpenClaw, an open-source autonomous AI agent platform.
- Q4 2025: Venice V2 platform rollout brought video generation, credit system, and deeper VVV integration.
Competitive Landscape
Venice sits in the AI inference layer, but it is not a decentralized compute network. It is a privacy proxy that aggregates existing models.
- Bittensor (TAO): A decentralized network where models compete to produce intelligence, with a $1.84B market cap. Bittensor is compute infrastructure; Venice is a consumer application layer. Our Bittensor analysis covers the network in detail.
- Render Network (RNDR): Decentralized GPU compute for rendering and AI. Render provides raw compute; Venice provides inference access.
- Surplus Intelligence: An open order book for AI inference where sellers compete on price. Venice routes to fixed providers; Surplus runs a live auction.
Venice’s closest competitor is not a crypto project. It is OpenAI’s ChatGPT and Anthropic’s Claude. The difference is privacy and censorship. Whether that difference matters to enough paying users is the core bet.
Strengths and Risks
Strengths
- Real product with 4 million users and verifiable API volume, not a whitepaper.
- Founder with a proven crypto track record and a named, experienced CTO.
- Aggressive emissions reduction (86% cut from launch) with on-chain vesting transparency.
- Programmatic burn mechanics linking platform revenue to token deflation.
- Four-tier privacy architecture gives users genuine choice over privacy trade-offs.
Risks
- VVV is currently net inflationary. Combined ongoing burns of ~31,833 VVV/month are far below the ~571,630 VVV/month in emissions and vesting.
- Venice is a centralized company, not a decentralized protocol. The team controls the treasury (part of its 35% company grant), the incentive fund, and the DIEM supply target. VVV carries no governance rights.
- Closed models (Claude, GPT, Gemini) still receive prompt text. The “private” branding is accurate only for Venice-hosted models, which are less capable than frontier closed models.
- The 67% FDV-to-market-cap gap means significant token overhang from treasury and vesting supply.
- The seven-day cooldown for unstaking sVVV limits liquidity and creates exit friction.
Analysis and Outlook
Venice is one of the few crypto-AI projects with a product people actually use and revenue they can prove. The $65M Series A at a $1B valuation validates that outside investors see a real business. The progressive emissions cuts and programmatic burn engine show a team that understands tokenomics.
But the deflationary VVV thesis remains unproven. At current burn rates, it will take years for revenue-funded burns to outpace emissions. The January 2027 vesting cliff helps by removing the biggest source of supply pressure, but the burn budget needs to grow substantially. That growth depends on whether enough people care about private, uncensored AI to pay for it.
The product works. The privacy architecture is genuine for hosted models. The token mechanics are transparent and verifiable on-chain. The team is doxxed and experienced. But VVV at a $1.16B circulating market cap prices in a lot of future success. The token is a hold for existing holders who believe Venice can grow its way into deflation. New entrants should size positions with the understanding that the deflationary crossover is not guaranteed and may be years away.
Sources
- Venice.ai VVV Token Page
- Venice.ai Series A Announcement
- VeniceStats Tokenomics Dashboard
- Venice Development Update October 2025
- Programmatic VVV Buy and Burn
- Tokenomics Update: Credit Burns and DIEM Supply Expansion
- CoinGecko Venice Token
- Venice AI on IQ.wiki
This article was drafted by agentbhm, an AI research assistant that does the legwork so the human editor can focus on judgment calls. Think of it as a caffeinated intern who reads whitepapers at 3am but still needs someone to double-check the math.