Tony Dinh: How He Built a One-Person Business

by TFE Editor · 11 min read
Tony Dinh: How He Built a One-Person Business

Tony Dinh is a software engineer from Vietnam. He wrote that 20 September 2021 was his last working day, and that he was working in Singapore. He sold Black Magic and Xnapper, hired help, and wrote that TypingMind revenue passed $1 million in the twelve months to November 2024.

A solopreneur starts and runs a business without a standing staff. His pages put a freelancer on TypingMind in April 2023, a full-time hire and three freelance developers by September 2023, and “my team” in June 2026. A freestyle entrepreneur builds the business around a life. Every dollar below is dated and attributed. Figures below are his, fetched November 14, 2026, and unaudited.

Tony Dinh at a glance

Item What the sources support
Known for TypingMind, DevUtils, Black Magic (sold), Xnapper (sold), a monthly newsletter, and a book he called MyIndieBook
Model One-time licenses where the app has no server, subscriptions where it does, two documented sales, and an audience he writes to
Team Solo founder of Black Magic, in his June 5, 2023 introduction. A freelancer by April 2023, then one employee and three freelancers. June 2026: “my team.” The about page fetched November 14, 2026 gives no headcount
Notable work TypingMind, personal and team. DevUtils. Black Magic, sold to Hypefury. Xnapper, sold to a buyer he names as Brian. Midday, an email client, with a December beta planned on November 27, 2025. Survive 10 Waves, a free game, on August 29, 2026
Where to follow news.tonydinh.com. On November 14, 2026 that page said “Over 15,000 subscribers.” The April 2, 2023 post shows @tdinh_me. TypingMind is at typingmind.com. www.tonydinh.com redirects to that newsletter

The story, from a failed Mac app to TypingMind

Six months, then two weeks

On September 22, 2023 he published the two-year recap. In 2021 he was a software engineer with 7 years of experience. Vietnam is his hometown. Early 2020 he worked remotely in a foreign country and listened to Indie Hackers, including Pieter Levels. An unnamed Mac log viewer took about six months, then he stopped. DevUtils took about two weeks: an offline toolbox, free to people he knew, then $9 once. Hacker News put it on the front page. The first dollar came in Singapore, while he was employed. A week after Product Hunt, he rarely got visitors. The $9 price was a one-time purchase.

The last day at the job

From 100 followers in November 2020 he reached 700 by May 2021. Near 1,000 he turned a profile-picture progress bar into Black Magic and charged $4 a month. August 2021: about $300 in monthly recurring revenue, about $200 a month from DevUtils, about 8,000 followers, and about 1,500 active users, most free. He was still employed, in Singapore. He wrote that 20 September 2021 was his last working day and his first day as a full-time indie hacker. He wrote that he had two years of savings, and that living in Vietnam could stretch that to four years with no revenue. The goal was $1,000 in monthly recurring revenue in year one. The newsletter, started around then, showed 6,000-plus subscribers in that post and over 15,000 on November 14, 2026.

I’ve always prefer to go solo.

Tony Dinh, September 22, 2023. The same post says he started a team in the second year anyway. He hired one full-time employee, for content, marketing, and support, and three freelance developers. More founder pages are on the solopreneurs hub.

The API bill, then two sales

After the job he worked about 12 hours a day, or 16 if Twitter counted, then about 4. The Magic Sidebar became the feature he rebuilt the product around. February 2022: $4,000 in monthly recurring revenue and 28,000 followers. October 2022: Black Magic at $13,000 in monthly recurring revenue. Xnapper made $6,000 a month by September 22, 2023. On April 2, 2023 he wrote that EmojiAI and AskCommand made about $100 together.

Around February 2023 Twitter announced it would end the free API. On April 2 he put the minimum at $42,000 a month and Black Magic at about $16,000 in monthly recurring revenue. A free plan and a $100 plan, he wrote, were too limited to run a serious business. On June 5, 2023 he named the price.

I sold Black Magic for USD$128,000

That is the total. About $14,000 a month at the sale, or $168,000 in annual recurring revenue. He had rejected $500,000 about three months earlier. A March contact had said there was no smaller plan. The April 2 post says otherwise. Three buyers, two offers, Hypefury, escrow.com. The API was cut around March 30 and then restored. A $5,000 plan arrived after the close. He would have paid that price a month earlier, and he did not check the new plan. He would keep working on Black Magic for at least another year. The introduction calls him the solo founder.

TypingMind

The ChatGPT API arrived on March 1, 2023. He registered typingmind.com on March 2 and shipped on March 6. By March 10 he had $10,000 in revenue.

99% of the sales are from my Twitter reach.

March 11 was Product Hunt’s number one product of the day. The September 22, 2023 recap says seven days came to $22,000 of license revenue. A March 12 tweet in the April 2 post prints $22.7K total. He stopped sharing the revenue after the launch.

No backend, no server, and no database.

Customers stored their own key, a self-hosted copy existed, and he charged once. By September 22, 2023 the price had moved from $9 to $39. That is not a 2026 price.

It’s no longer a one-man show

Tony Dinh, April 2, 2023, when a freelancer had started on TypingMind. He also wrote, “This newsletter is where I can control my own destiny.” On September 22, 2023, TypingMind was about $30,000 a month, inside about $45,000 a month at about 90 percent profit. His last drawn salary was about $9,000 a month. That product is the small-software shape of micro SaaS.

February 26, 2024: $500,000 total and $15,000 in subscription monthly recurring revenue. March 19, 2024: Xnapper sold for $150,000, $100,000 upfront and $50,000 as a revenue share, while TypingMind made $40,000-plus a month. Brian, the buyer, came from a tweet. A $169,000 offer arrived after he accepted. He kept DevUtils and TypingMind. November 22, 2024: $1 million in revenue in the last 12 months, one 3,000-seat deal and several 1,000-seat deals, and at least $20,000 on compliance, including SOC 2, HIPAA, GDPR, security audits, and lawyers. He wrote that TypingMind was SOC 2 Type II, HIPAA, and GDPR compliant. The about page fetched November 14, 2026 does not say so.

What later posts do not update

On June 15, 2025 he took book pre-orders at $24 and scheduled July. Chapters were unfinished, and no page fetched here gives a ship date. On November 27, 2025 a Black Friday line listed MyIndieBook at $9, and he planned a December beta for Midday, an email client. The June 10 and August 29, 2026 posts do not mention Midday, and they print no new monthly revenue. The Growth Reads sitemaps had no review of the book.

On October 9, 2025 TypingMind was about $130,000 to $160,000 a month, team recurring had passed half, and $600 of ads made 2 conversions. That post says he joined David, Travis, and Phuc to plan a one-month residency for 10 indie hackers in Da Nang in November 2025, mostly led by Travis. On November 27, 2025 the demo day was November 29. On June 10, 2026 he called the November event co-hosting, and a second batch ran in May 2026. On June 10, 2026 he relaunched TypingMind on Product Hunt. “My team and I” would be on the page. He wrote that he gave Claude Code full permission over a Mac Mini after waiting because of prompt injection.

I didn’t write or read any code.

Tony Dinh, August 29, 2026, about the free game Survive 10 Waves. He also wrote that he was still writing code every day, in the October 9, 2025 post. The homepage fetched November 14, 2026 says both “20,641+ happy customers use TypingMind daily” and “hundreds of thousands of productive users using TypingMind daily.” The Custom about page says the team has grown significantly. Its counters are undated.

The playbook: five moves he actually dated

1. Time-box the first version

Six months on a log viewer, then he stopped. Two weeks for DevUtils. Five days from the API post to TypingMind. He wrote the rule on September 22, 2023.

Ship early, ship small, ship frequently.

2. Keep a list you can still send

Twitter carried the early TypingMind sales. By October 9, 2025 most sales were word of mouth, and $600 in ads made 2 conversions. The newsletter is the channel he said he controlled. We like Kit for this: it’s free up to 10,000 subscribers. Partner link · T&Cs apply. Kit’s help center, updated September 23, 2026 and fetched November 14, 2026, says “The Free plan costs $0 per month” “for as long as you stay under 10,000 subscribers.” It includes “Up to 10,000 email subscribers” and “Unlimited email broadcasts.” Visual automations, email sequences, and AI Connectors sit on paid plans. Past the cap, the help center says “You’ll need to upgrade to a paid plan to keep sending.” The pricing page fetched the same day shows Free at “$0” “/mo” and “Free—up to 10,000 subscribers,” “up to 10,000 subscribers, for free,” no sending limits on any plan, and an automatic upgrade that keeps sending. Past 10,000 the pages disagree. A slider reading “For 1,000 email subscribers” is not the cap. His list is on Substack. Nothing fetched here says he uses Kit.

3. Match the price to the server

He charged once while the app had no server, and he added subscriptions when sync and the team product needed one. By October 9, 2025 the team subscription was the larger half.

4. Write the month a platform can erase the product

Put the API price next to that month’s revenue, and split a sale into cash now and cash later. His $128,000 and $150,000 headlines both need the terms beside them.

5. Say who else is in the room

Call it solo only for the years the page still supports. His own posts add a freelancer in April 2023, an employee and three freelancers by September 2023, and “my team” on June 10, 2026.

What did not work, and the caveats

  • Two early products barely paid. The log viewer has no name and no price. On April 2, 2023, EmojiAI and AskCommand made about $100 together.
  • Solo is dated. The June 5, 2023 introduction says it about Black Magic. By September 22, 2023 he had hired, and June 10, 2026 still says “my team.”
  • The monthly figures move. Black Magic is about $16,000 in monthly recurring revenue on April 2, 2023, about $14,000 in monthly recurring revenue at the June 5 sale, and $13,000 in monthly recurring revenue in October 2022. Xnapper is $6,000 a month on September 22, 2023 and about $4,000 a month at the March 19, 2024 sale. The $150,000 was not all cash up front, and a $169,000 offer came too late.
  • TypingMind is several metrics. Seven-day revenue, a monthly rate, a trailing twelve months, and the October 9, 2025 range are not one number. The April 2 title uses $22,000. A March 12 tweet in that post prints $22.7K total. Later posts do not update the month. The homepage disagrees with itself on customer count. The $39 price is the September 22, 2023 “now,” and “60% on the premium license” is the November 27, 2025 Black Friday discount, not a price and not revenue.
  • The book, Midday, and the game need the date on them. July 2025 was a schedule, not a fetched ship date. Midday’s beta was a December 2025 plan. “I didn’t write or read any code” is the August 29, 2026 game, not the October 9, 2025 line that he was still writing code every day.

The AI version of this playbook

This week is ours. On June 10 and August 29, 2026 he used a coding assistant, and he still chose the offer. A tool drafts. You check it against a page you opened. That is AI for solopreneurs.

Job You decide AI drafts You check
The first version The chore and the deadline A one-job description It matches a real annoyance
The price Once, or a subscription A sentence from the cost you typed No hidden bill under a lifetime deal
The number Total, month, or forecast A sentence from figures you typed Cash received is not the headline
The names Who else builds or owns An About line with every name The page shows them
The list What you send, and when A draft from your outline You send it yourself

A first week, if you already know the chore. This plan is our analysis.

  1. Day 1. Write the chore and the hour it costs.
  2. Day 2. Cut it to a two-week version.
  3. Day 3. Price it once, or as a subscription, from the real cost.
  4. Day 4. Send one note where people already talk, and keep a copy on your list.
  5. Day 5. Name everyone else on the page.
  6. Day 6. Publish one number, labeled revenue, cost, or a forecast.
  7. Day 7. Check the draft against a page you opened.

Lessons for your one-person business

  • Ship on a clock. Six months failed. Two weeks and five days shipped.
  • Keep a newsletter you still control. He said that on April 2, 2023, then sold Black Magic when the API price did not fit.
  • Print the terms. Xnapper’s $150,000 included $50,000 of revenue share. The $128,000 sale followed a rejected $500,000.
  • Name the other people. A freelancer is on the April 2, 2023 page. “My team” is on the June 10, 2026 page.

FAQ

Who is Tony Dinh?

A software engineer from Vietnam. The September 22, 2023 post gives him 7 years of experience as of 2021, work in Singapore, and a last working day of 20 September 2021. He sold Black Magic and Xnapper and still writes the newsletter, which showed over 15,000 subscribers on November 14, 2026.

Did he build TypingMind alone?

He shipped the March 2023 version himself. A freelancer had joined by April 2, 2023. By September 22, 2023 there was one employee and three freelance developers. June 10, 2026 says “my team.” The about page gives no headcount.

How much money has he reported?

His numbers, unaudited. August 2021: about $300 and about $200 a month. October 2022: Black Magic at $13,000 in monthly recurring revenue. April 2, 2023: the title’s $22,000 of TypingMind revenue in seven days, a March 12 tweet at $22.7K total, and Black Magic at about $16,000 a month. June 5, 2023: a $128,000 sale at about $14,000 a month. September 22, 2023: about $45,000 a month across products. March 19, 2024: Xnapper for $150,000, with $100,000 upfront. November 22, 2024: $1 million in revenue in the last 12 months. October 9, 2025: about $130,000 to $160,000 a month. The June 10 and August 29, 2026 posts do not update that month.

What happened to Black Magic and Xnapper?

Hypefury bought Black Magic. The June 5, 2023 post prints $128,000 after a rejected $500,000 offer, and says he would keep working on it for at least a year. He sold Xnapper on March 19, 2024 for $150,000, $100,000 upfront and $50,000 as a revenue share, to a buyer he names as Brian. He kept DevUtils and TypingMind.

Does he have a book or a newsletter?

The newsletter is the Substack at news.tonydinh.com. He reported 6,000-plus subscribers on September 22, 2023, and the homepage said over 15,000 on November 14, 2026. He scheduled a book for July 2025 at a $24 pre-order, then a Black Friday line listed MyIndieBook at $9 on November 27, 2025. No fetched page gives the ship day, and The Growth Reads sitemaps had no review.

The takeaway

Copy the public record: a small version on a clock, a price that matches the server, a list you still send, and every other name on the page. When the number changes, date the new one.

Grab the free One-Person Business AI Stack checklist and use it to staff that week: one number you typed, the other names, and a list you still send.

Sources

Fetched November 14, 2026, unless the date is the page’s own.