Loading page…
One subscription that unlocks every top LLM plus image and video models in a single chat console — MVP built on no-code Bubble in 8 weeks, scaled to nearly $100K/mo on a decade-old personal brand and a recurring-commission affiliate program
The real moat isn't the product — it's the 100K-subscriber email list a decade of building in public produced: buyers were already waiting on launch day, so no ads were needed. And a non-technical founder can ship a SaaS people pay for on Bubble in 8 weeks.
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
The product is Magai (the Chinese auto-subtitles garble the name as Magi/MAGI/Matai — the correct spelling is Magai), a unified chat console that bundles every top-tier LLM (GPT-5, Claude, etc.) plus image/video models (Dall·E, Flux, Ideogram, etc.) into a single $20/mo subscription. The pitch: “one subscription replaces 5-6 AI subscriptions.” Founder Dustin Riechmann has no technical background and previously ran a WordPress plugin company; he was interviewed on the Starter Story channel hosted by Pat Walls.
There was no cold start, and Dustin is direct about why: the most effective marketing he ever did was the personal brand he spent the previous decade building. A blog he had run for ten years had gathered more than 100,000 email subscribers, so on launch day the buyers already existed and no advertising or traditional marketing was needed. His rule for how that list got built matters more than the list itself — show your progress openly and transparently, including the parts that aren't polished and look nothing like a curated feed, because people want to watch the process. Two other channels ran alongside it. He is an extrovert who invested years in genuine relationships across the marketing industry, and friends there, strong marketers themselves, promoted the product. And the affiliate program was wired into the product from the beginning rather than bolted on later. None of that requires a paid channel, which is how a $20-a-month tool from a non-technical solo founder reached $3,000 in its first month.
He shipped expecting nobody to buy, and the first customers arrived the day it went live: $3,000 in month one, $4,000 in month two. There is no free tier and no trial. Magai is paid from the first click at $20 a month for an individual, covering an average month's usage with more available to purchase on top, and $40 a month for a team of up to five sharing roughly three times that allowance, with additional seats at $20 each. The pricing does two jobs at once. It turns the pitch into arithmetic rather than adjectives — someone paying for five or six separate subscriptions to get the best of every model collapses them into one $20 line, and the saving is calculated before any feature list is read. It also protects a business with a real cost of goods, since every conversation spends tokens through OpenRouter or Fal, so a free tier would have been a bill rather than a funnel. Topping up usage means the heaviest users pay for themselves.
The best-evidenced and most genuinely business-like of the five: a named founder, a clear revenue curve, and a credible stack (Bubble + OpenRouter + Fal + Stripe + ConvertKit). But don't let the “8-week build” mislead you — what actually makes the money is the decade-old audience and the recurring affiliate program, and that's the part you can't copy. If you want to learn from this, learn “build in public, accumulate an audience,” not “aggregate a bunch of models.”
Paste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself.
Data credibility: ⚠️ Revenue is self-reported by founder Dustin in his interview on Pat Walls' Starter Story channel ($3K month one → $4K month two → $10K month three → $30K/mo at one year → nearly $100K/mo now). The show includes a product demo but no third-party financial verification, and that episode is sponsored by Rocket (rocket.new). This is secondhand interview reporting. All revenue figures are founder-reported or estimated by the video creator — unaudited.
The affiliate program is what made growth self-propelling, and the design is the reason. Commission is recurring rather than a one-time bounty: if a customer pays twelve times, the person who referred them is paid twelve times. That changes who bothers to promote it. A single payout attracts people hunting a launch spike; a trailing commission attracts creators and marketers willing to keep mentioning the tool a year later, because their income now depends on those customers staying subscribed. Paired with a product that is hard to cancel — one $20 subscription standing in for five or six — every promoter builds a book of business instead of collecting a payday. The curve shows where it caught: $3,000 in month one, $4,000 in month two, then $10,000 in month three, $30,000 a month at the one-year mark, and close to $100,000 a month now against more than $1 million cumulative. That third month is where a launch to an owned audience handed off to something that kept running without him.
The hard stretch came before Magai rather than during it. His first software venture, a WordPress plugin company started in 2014 with partners, worked almost without effort; the partners drifted apart, he left, and the two ideas he tried next failed outright and never launched. He describes being lost after that, genuinely unsure whether he still had the ability, and moving forward anyway. The lesson he pulls out is about interpretation rather than endurance: the biggest thing founders get wrong is letting past failures define them, because when people rejected those products they were rejecting the products and nothing more. What he changed on the next attempt was the size of the bet. With no money to hire a developer, he found an online course, learned as fast as he could, and had a complete MVP running on Bubble plus a little code he understood within eight weeks, with Stripe attached so the verdict would arrive as a payment. Eight weeks is short enough that being wrong costs a season, not a year.
Self-reported numbers: $3K in month one, $4K in month two, $10K in month three, $30K/mo at the one-year mark, and now close to $100K/mo, with over $1M cumulative. Pricing: $20/mo for individuals; $40/mo for teams including 5 seats, then $20/mo per additional seat. Paid-only — no freemium tier.
Front end built on Bubble, no code — MVP shipped in 8 weeks. OpenRouter as the single gateway for LLMs, Fal for image/video/audio, and a handful of models wired straight to their own APIs. Stripe for payments, ConvertKit for email, GA + Ahrefs for analytics. Three growth pillars: a decade-old personal blog with 100K email subscribers, a recurring-commission affiliate program, and endorsements from his industry network — with almost zero ad spend.
This episode is sponsored by the no-code builder Rocket (rocket.new), so the narrative is engineered around the “you could build this in 8 weeks too” hook, and the revenue figures are self-reported with no third-party verification — discount them accordingly. That said, compared with the pure lead-gen channels, this case has a named founder, a clear revenue curve and a credible stack: it's the most real-business-looking of the five. The transferable lesson is “build in public to accumulate an owned audience,” not the model-aggregation idea itself.
That was one of the 50 breakdowns open to everyone. The Unicorn Club opens the other 1,285 at the same depth — plus pain-point filters and 10 build specs a week.
Join the Unicorn Club — $5/moCancel anytime, in one click