Dappit takes someone with no code from an idea to a live, audited, on-chain product — in one app. Images, sandboxes and deploys already run on GPUs we own, and we are opening them as a node network. Every other AI builder rents its whole stack and stops at the website.
Every other AI builder rents its whole stack from a foundation lab and passes the bill to the user as a subscription. We own the hardware, the sandboxes, the assets and a model of our own — so a build costs us electricity, not API credits, and we only charge when you ship. 5,100 builders signed up · 40 builds a day.
Single-use AI is a feature — anyone with the wherewithal can rebuild any one thing we do. Nobody has combined them, and nobody paying a foundation lab per token can give the whole flow away for free. The user pays nothing until they have something worth shipping.
A competitor can clone any single row. To clone the column, they would have to build all eight — and own the model, the assets and the metal underneath them.
Two engineers who have shipped on-chain for years, and two of the most-watched interviewers in crypto. Every user so far arrived through Product Hunt, X and partner announcements — $0 of paid acquisition, and the founders' 250K-subscriber channels have not been pointed at the product yet. That is the launch asset, not the growth story to date.
6 years in blockchain. Two successful startups behind him. Owns the chain relationships and the BitAngels HK win.
Degree in robotics engineering. DeFi engineer — smart contracts and infrastructure. Builds and runs the sovereign fleet and the trained model.
Founder, Whale Coin Talk + Moby Media. 250K+ subscribers. Interviewed Charles Hoskinson and most of the majors. @mobymedia · @whalecointalk
7 years working directly with chains. Thousands of interviews shipped. Channel reach held for the $DAP launch. @player1taco
Most teams raise to buy distribution. We have not spent a dollar on it, and we have not spent the channels either — they are held for the launchpad and the token launch.
$DAP is not a token bolted onto a SaaS. It is the settlement unit of a sovereign compute network — burned when a user ships, earned when a node serves. One billion supply. Fixed. Demand tracks usage, not narrative.
The product stays free forever. The sovereign rails — deploy, DNS, web3, heavy GPU, custom models, token launches — settle in $DAP. The token is not the business model. It is the meter on a business that already works.
Insiders — team, advisors and both priced rounds — hold 40%; company-controlled buckets are 45%. Team carries the longest cliff: 12 months, then a 36-month vest. Two priced rounds at 10% each. Only a thin slice unlocks at launch.
A network of machines cannot be paid in equity. The moment compute leaves our own racks and runs on somebody else's hardware, that machine needs to be paid — per job, permissionlessly, in something it can hold. That is the entire reason $DAP exists.
The user burns the free tier, sees the wall, and leaves. We removed the tax by owning the model and the hardware. But owned hardware has a limit — ours. To go past it, other people have to run our model on their machines and get paid for it. That is a network, and a network needs a settlement unit. The token is how a thing you own pays for itself.
Third-party operators serve inference on their own GPUs. You cannot pay a stranger's machine in equity, or in invoices, or in trust. 10% of supply is reserved to pay them.
Subscriptions die on API costs. Usage settles on the rails instead — and every settlement burns supply.
This round buys the B200 genesis node. Tokens align every operator who joins after it with the fleet that started it.
The people who took themselves from zero to shipped built our traction. Points convert them into holders — not farmers who found us yesterday.
Dappit.fun settles launch fees in $DAP and pays stakers from them. The launchpad and the token are one product.
Owning the model, the metal and the deploy is expensive. A token treasury funds the thing that does not depend on anyone else's switch.
Two things back the token: a product that already ships, and lockups that keep us in the chair. The rails below are live today — not a roadmap item, not a testnet. And the people who built them cannot sell into you.
Long enough to filter farmers, short enough to reward builders. Team on a 12-month cliff and 36-month vest. Both priced rounds cliffed. Only 13.9% circulating at TGE.
Ten of those points are market-maker inventory, not sellable float. Everything else is cliffed, vested, or committee-released. All cliffs run from TGE. Insiders — team, advisors, and both priced rounds — total 40% of supply, and the last insider tokens release in month 48. The people who build the network are locked into its long arc.
Most tokens need a story to hold a price. This one needs traffic. Six hooks, one loop: every shipped build removes supply, every served job pays a node, and every node added lowers the cost of the next build.
Every deploy and heavy op partial-burns $DAP. Supply tightens as usage grows.
Hosting, DNS, web3, GPU, custom models — the paid rails settle in $DAP.
Operators run the Dappit model and earn $DAP from a dedicated 100M-token bucket, released over 48 months against verified compute served. The genesis fleet does not earn from it.
Staking unlocks priority access and higher limits on Dappit.fun launches, subject to eligibility and jurisdiction.
Launchpad fee revenue funds fee rebates and priority compute — usage, not emissions. Final mechanism subject to legal review.
Stake for fee rebates, priority compute, and a vote over the treasury.
The only token model that survives a bear market is one where the buyer is a user, not a speculator. Ours is metered by product usage.
The technical environment already exists to carry a user from an idea and no code to a shipped, funded, tradable product — without ever leaving Dappit. Every rung of that climb is a place we get paid, and the user pays nothing until the moment they have something worth shipping.
An idea, no code, no team, no budget. The wall where every other platform asks for a card.
Prompt → full stack. Frontend, contract, tests. Twelve art-directed worlds so it doesn't look AI-made.
Deploy to the node network, custom domain, hosting. The first moment it is real.
Wallet, smart contract, audit, gas. The part every AI builder stops short of.
Token on Dappit.fun with a real product behind it. Bonding curve, chart, holders, the directory page.
Custom model, custom personality, 3D, video, heavy GPU, enterprise node. The winners spend here.
A live product, a funded community, a tradable asset — and a builder who never had to leave, or learn to code.
Nobody else owns the whole climb. Bolt and Lovable stop at rung 01. Pump.fun starts at rung 04 and has no product underneath it. We own every rung — and $DAP is the meter on all of them.
$4M pre-seed across two tranches. Each round buys 10% of the token supply and 10% of the company — Round 1 at a $10M valuation, Round 2 at $30M. Round 1 is open now at the best terms. Sarson Funds leads.
10% of the token supply + 10% of the company, at a $10M valuation. Best-terms entry — the genesis round that funds the B200. One-month token cliff.
10% of the token supply + 10% of the company, at a $30M valuation. Scales the node network into mainnet.
Every round is 10% of $DAP and 10% of the company at the same valuation. Equity VCs get equity; crypto-native VCs get token upside too.
B200 genesis node · model training · engineering · audits · liquidity + MM · listings · GTM. Twenty-four months of runway.
Lead investor committed. Chain — Solana, ETH L2, or both — optimized with the lead. Round closing.
Public beta v1 launched July 2025. Thirteen months later: 5,100 signed up, 40 builds a day, $800 MRR, four chain partnerships, and an award from Hong Kong. The thesis is validated. V3 is the unlock.
Every build pulls from databases we own. The spine compounds with every user — and no competitor can build it after the fact.
Thirteen months from 0 users → 5,100, from public beta v1 to first revenue. Featured on BNB Chain. Hackathon with Neura. Winner at BitAngels Hong Kong. Partnership with LitVM. Countless grants — NVIDIA Inception, AWS Activate, and more. The next 24 months scale the rails, not the bet.
Builder + media. Product Hunt top 14. ~300 users.
Wallet, contract scaffolds. Smart contract audit gen in alpha.
Image gen · Design Studio · pump.fun deploy · smart contract audits. ~800 reg · ~50 DAU.
Dappit announced as a featured new app on BNB Chain.
Partnership with ANKR's new EVM chain. Builder integration.
Win the inaugural BitAngels Hong Kong pitch. LitVM partnership signed.
Sovereign deploy live. Pricing live — $800 MRR. 5,100 signed up · 40 builds/day.
5,100 people signed up. 40 build every day. Almost none subscribe. Not because the product is wrong — because the economics underneath every AI builder in this category are broken. We hit the same wall Bolt, Lovable, v0 will hit. V3 removes the wall by removing the tax.
Every dollar that used to flow to a foundation model API now flows to our own hardware, our own model, and the DAPPIT node network. Marginal cost approaches zero. The product becomes free. Revenue moves to the rails we control.
Custom-trained Dappit model. No per-token API rental.
Sovereign fleet. No cloud rental, no AWS bill.
DAPPIT node network. Marginal cost → zero as nodes join.
Kill the subscription wall. Build for free, forever.
Deploy, DNS, web3, 3D, custom models, token launches.
AI inference demand is hot. Idle GPUs rent to the open market for additional revenue — the asset earns even at rest.
The next billion-dollar product in this category is not the one with the smartest model. It is the one that cannot be turned off — by a foundation lab, by a cloud, by a payment processor, by a UX pivot — and the one that does exactly what the user needs. Claude is an expert at everything. The Dappit model is an expert at Dappit, your company, your workflow, or any subject you choose.
Every other AI builder ships visibly AI-generated output. Dappit's model executes a look — it does not improvise one. Twelve art-directed worlds, each with a prescriptive styleBrief. The spine — every image, every 3D, every template, every render — learns from every build. The next user inherits the gain.
The 12 worlds are not static templates. They are living patterns powered by a sovereign database we own. Every build deposits — images, 3D, copy, templates, knowledge. Every render teaches the model what works in context. The next user inherits the gain. No competitor has this loop.
20,241+ images · 124+ GLBs · 12 worlds · §B0–§B10 knowledge — all owned, all growing, all served live from assets.dappit.app.
Every build deposits images, copy, templates, 3D, video into the catalog spine — CLIP-embedded, aesthetic-ranked, curated, dedup'd. The longer Dappit runs, the cheaper, faster, and more beautiful every new build becomes. No competitor has this. None can build it after the fact.
Getty charges $30 per image. Shutterstock charges $99/month. Adobe Stock charges $80. Every other AI builder is either paying these tolls — or building on assets they cannot legally license. Dappit pulls from a sovereign spine that is license-clean by source and AI-augmented by demand. AI has no trademark. AI has no IP. Today we save our users the tax. Tomorrow we open the spine as an API and collect it.
US Copyright Office, 2023: AI-generated images are not human-authored, so no individual or corporate IP attaches. Public domain by default. The lawsuits the labs are fighting? We sit downstream of all of them.
Shutterstock's revenue is flat. Getty is in court. The $5B/year licensing market saw the cliff and trained their own models too late. Their assets are not improving. Ours compound with every build.
Every Dappit build pulls from a sovereign, license-clean, AI-augmented spine. Tomorrow, other AI builders pay us to use it. The tax base becomes the toll booth.
The asset spine is not just our moat. It is a B2B API product the market is already paying $5B/year for. Every Bolt clone, every Lovable successor, every AI agency that ships in 2027 needs license-clean assets, on demand, at scale. assets.dappit.app opens to all of them — at a price they cannot beat, because nobody else can build what we already shipped.
Dappit users build with the spine for free. Switching cost compounds with every render they save from Getty.
assets.dappit.app opens to AI builders, indie devs, agencies. Per-pull billing in DAPPIT or fiat. We undercut Getty by an order of magnitude and still print.
Custom-curated spines licensed to platforms with revenue share. The asset CDN as sovereign infrastructure — the Cloudflare of AI-clean content.
Investors fund the NVIDIA B200 genesis server — the inaugural node. The Dappit fleet bootstraps the network. Third-party operators run the Dappit custom-trained model and the media pipelines on their own hardware and earn DAPPIT for serving compute. When the network opens, they cannot 51% us. We are the genesis.
The product is free. We monetize the sovereign rails — hosting, DNS, web3, heavy GPU work, custom personalities, custom models. B2C unlock: best builder on the market, free. B2B unlock: private enterprise nodes, sovereignty as a sales pitch.
Pump.fun proved fair-launch tokens print money (~$700M in 2025). Dexscreener proved tier-gated information prints money. Neither ships a site. Dappit is the only product that does all three — and the brand makes Web3 look like grown-ups.
Bonding curve · virtual liquidity · fair start. Token live in minutes, free wildcard subdomain assigned.
Free entryDexscreener-grade chart, holders, volume, social links — public, indexable, embeddable. Becomes the directory.
Drives organic trafficThe winners pay for a real site, custom domain, full design world, wallet + contract rails. Built in minutes.
→ Web3-exclusive revenueSix quarters. Video studio → token whitepaper → node software → token launch → daydream agents → public launch → enterprise nodes. The B200 commissions in Q3 2026. Mainnet is Q1 2027.
LTX 2.3 live, 52s/clip. DAPPIT token whitepaper. B200 commissioned.
Node operator software alpha. spec→deploy. 3D auto-rig. Gallery + discovery.
TGE. Third-party node onboarding. Game asset forge. LoRA as a feature.
AI employees. Per-build company checklists. Build companies, not websites.
Global signup. Pay-per-deploy in DAPPIT or fiat. 100k+ image catalog.
Private enterprise node SKU. 100+ third-party nodes targeted.
Vertical sovereignty plus decentralized hosting plus token-aligned compute equals a moat that no incumbent can close. When the market consolidates in 2028, the survivor is the one who owned the road and put the road on a network.
Confidential — for discussion with prospective investors only. This document is not an offer to sell or a solicitation to buy any security or digital asset, and is not investment advice. Token utility, allocations, vesting and timelines are indicative and subject to change, to legal review and to lead-investor input. Forward-looking statements are estimates, not commitments. Any offering will be made solely to eligible investors under definitive documents and applicable securities law. Metrics are as of 3 August 2026 and unaudited.