Would you consider a pay-as-you go option, perhaps with a deposit?
Right now, I won't subscribe for $12 per month (or $25 for API). But I have some use for both the interactive and API conversions and am happy to pay per conversion, so a pre-paid version (like OpenRouter) would be great.
3 hours ago [-]
esFran 4 hours ago [-]
It’s a beautiful story. Having a business like this in the AI era is VERY inspiring. I see you as a small business, like the butcher or the fishmonger, selling good products to the public. Congratulations, and keep it up.
muhuhuhaha 14 hours ago [-]
Are you noticing much of a hit from people using claude code/ desktop etc to do conversions?
theanonymousone 1 days ago [-]
Really happy for you. It's my dream to run such a business for so long, as a side, with a real user base, and not for money.
CalRobert 1 days ago [-]
How is it a business if not for money?
throwaway83772 1 days ago [-]
I interpreted it as "enough to live a comfortable life".
Money is part of how you get there, but it is not a goal in itself. Many start businesses where money is the main goal, and where the amount that you are satisfied with inevitably shifts once you get richer
teiferer 1 days ago [-]
Considering that their cheapest pricing is $12 (per month, 1 month minimum) and they claim 550 million users, either they made tons of riches or the numbers are to be questioned. I leave it to you to estimate which is more likely.
andai 1 days ago [-]
I don't think that means paid users.
They also have a paid API.
weird-eye-issue 22 hours ago [-]
Is this your first time encountering a freemium business model?
j45 20 hours ago [-]
Doubt worshipping finds ways to not find a way to make something happen. Just because one person can’t see it doesn’t mean it’s not possible.
theanonymousone 1 days ago [-]
Let's say money as a proxy for "being non-trivial" or worth enough for someone to pay for. I don't know how else this can be signalled.
"Side" in the sense that my main income coming from somewhere else, aka my "normal" job
j45 20 hours ago [-]
Time isn’t money, money is time.
Money that doesn’t need to be spent is freedom of time.
andsoitis 18 hours ago [-]
Money is valuable partly because it can buy time, but time is valuable independently of money.
j45 15 hours ago [-]
In a world where people trade time for money, they effectively collect time as money, and money can be used to buy back time through the correct conveniences if they choose to buy back the freedom in their time, first through high leverage low cost items, and then over time more.
kubbity 1 days ago [-]
Hi. Congratulations and thanks; it just happened that today I needed an .mp3 to .txt conversion and used your site. Nice to see it here now. All the best!
timcobb 1 days ago [-]
> I needed an .mp3 to .txt conversion
?
guydoom 1 days ago [-]
Also known as transcription, I suppose.
andai 1 days ago [-]
Does it give you the tabs?
guydoom 20 hours ago [-]
Only speech transcription, probably.
voidmain0001 23 hours ago [-]
Their web page provides a list of ‘convert mp3 to’ formats
Wow I have used this a long time ago and happy to find the origin story on HN.
Great job and thanks!
piker 1 days ago [-]
How has revenue changed during the AI period? Are you guys able to maintain or grow with the rise of coding agents?
[edit: what’s with the downvotes? I love the story and am curious if Claude etc have affected their business.]
f311a 1 days ago [-]
I’m pretty sure more than 90 percent of their audience never heard of Claude.
They are regular users, not developers.
FearNotDaniel 1 days ago [-]
Respectfully, it sounds as if your assumptions about the market for LLM/Agent tools is out of date by a long way. I don’t have any stats but it’s easy to observe that Claude, Gemini, ChatGPT et al are aggressively marketing their products to non developer business users, and pretty much everyone I know who is non technical has heard of at least one of those and uses it regularly in their work.
adastra22 1 days ago [-]
They said Claude. I'm pretty sure most normies have never heard of Claude. ChatGPT, yes. It's shortened to just "chat" now. But most people have no idea what Claude is.
martyvis 22 hours ago [-]
Claude etc have been on mainstream national news here in Australia multiple time in the last few months
andsoitis 18 hours ago [-]
ChatGPT can also convert between files.
Jtarii 23 hours ago [-]
>I don’t have any stats
ok.
embedding-shape 20 hours ago [-]
> pretty much everyone I know who is non technical has heard of at least one of those and uses it regularly in their work
Rather than using this as your world-view, I think you should take this as a signal to expand the current bubble you live in. I'm guessing most of these people you refer to as "non technical" are still white-collar workers or similar?
Most people in the world, even with all the marketing towards "non developer business users", has never heard of "Claude", "Gemini", "Grok" or anything else LLM-adjacent like that, but will have some light understanding of "ChatGPT", and might actually be using some AI service via their phone or whatever, but still call that "ChatGPT" as that's essentially synonymous with "AI" in common parlance.
Yeah, for example there are dozens of vibecoded file conversion tools being launched on Product Hunt every day. Most are free to use and supposedly work in pure js and files never leave your browser.
This is a case study of how a software company can still make money in the age of ai.
My guess is that the impact of ai on the business won't be too much as they have a good brand and a loyal customer base whereas the vibecoded ones will have a hard time getting noticed by customers. They will also have a competitive advantage by being able to handle so many different types of files, by being a one-stop-shop if you like. They will be able to attract more customers by handling really obscure file formats and in some cases by being able to do a better job of the conversion (desktop publishing and vector formats are an example). There is probably a power advantage of being able to convert n filetypes to m different formats where n and m are very large.
lelanthran 15 hours ago [-]
> There is probably a power advantage of being able to convert n filetypes to m different formats where n and m are very large.
If they've been in business this long, they must have, by now, internal filetypes for specific classes of formats. For example, all rgb files first convert to an RGB format, and then to the destination format. Or, all transcription first convert to an internal format, and then to the destination format.
Not sure how you'd handle .ps or .pdf files in this way, though.
j45 20 hours ago [-]
Only a tiny amount of people use ai this productively or know its possible
Fripplebubby 14 hours ago [-]
There are two threats from AI to this kind of business: 1) that they lose their users because their users use AI themselves to do their conversions and 2) that they lose their users to a wave of similar, shinier, more featureful competitors that spring up because suddenly the suite of features that took years to build in the Before Times can be built with AI. The extra competition could be consequential. I wish them the best, but the task ahead may not be easy. In this post they talk about how their work is meaningful because they feel they are touching people's lives behind the work they do. Will the vibe-coded competitors feel the same? Are we better off now that their paid services can be had for too-cheap-to-meter?
andai 1 days ago [-]
I use this all the time, especially for epub to pdf.
Tried doing it with pandoc and the result was similar but less good. (I like it when the result is more good.)
It's amazing how many formats they support. It's kind of strange that this isn't built into the computer, though. Without Zamzar, the computer still lives in the pre-Babelfish era...
My first thought was "imagine the astonishing amount of private documents and PII you could collect and store if you ran something like this for free as a honeypot". People just blithely uploading any document from their personal data to convert to another format.
And if I would have to guess it's less technical users who aren't going to try something like opening a document in the latest Libreoffice (or gimp, etc) and then exporting it to another format, using local software on their computer, first.
oxedom 23 hours ago [-]
I know for a fact that these honeypots exist.
j45 20 hours ago [-]
Yet some people still use them.
Why don’t you create an alternative?
walrus01 8 hours ago [-]
Frank answer: I'm far too lazy, it's outside of my general skill set of what I could confidently develop (or oversee hired people to build an equivalent service), and running a consumer SaaS with $10 to $30 monthly subscription/services or ad supported sounds like like a new level of personal hell for me.
sevenzero 24 hours ago [-]
Well this is what people do with AI. Uploading their whole life with no second thought. But surely we can trust big tech, right?
hn_submit 21 hours ago [-]
Are you guys using CGI-BIN or some other technology?
JakeAl 13 hours ago [-]
try .scn files from SoftImage XSI. Need a way to convert them to a format that can be used in Blender or Maya/FBX without losing any of the structure. Autodesk basically told the old users to bugger off leaving them with Gigs of files they can't use anymore.
gosen 1 days ago [-]
Great and very motivating story, thank you for that!
beaker52 22 hours ago [-]
Judging by the nature of the site (file conversion) and comments — has HN been invaded by bots here?
keyle 21 hours ago [-]
It is a little over the top and something is off with some of these accounts. Maybe there is a market to get yourself pushed to the front page of HN for money?
Group_B 19 hours ago [-]
Yeah this one looks very boted. Not typical HN comments at all
arbor-group 19 hours ago [-]
[flagged]
aitin 20 hours ago [-]
[flagged]
hamzaalibasra 23 hours ago [-]
[flagged]
jamiegodwin 1 days ago [-]
[dead]
umeshc7 23 hours ago [-]
[flagged]
Yash16 24 hours ago [-]
[dead]
teiferer 1 days ago [-]
Always glad to see success stories. But some stuff does not check out.
> Over 550 million people have used Zamzar since launch,
There are 8.3 billion people on earth. You are saying that about 1 in 42 have used your service? I had never heard of it and I don't know anybody who had. And that's a sample that should be biased in your favor. Sorry, does not pass the smell test.
Or maybe they were not unique visitors, containing duplicates? How many? This makes me skeptical of the remaining statistics on the page.
> converting over 766 million files. We now process over 1 million API requests per day.
1 million per day now and 766 million over 20 years? At the current rate you'd achieve that in about 2 years. This would indicate a rapid growth in the last couple years or even just months. It is unclear why that would be reasonable. Especially since you'd probably have emphasized this explosion more if your service was essentially dormant for 19 years and just recently went through the roof. (Sth like "how we handled 100x growth of our SaaS business over 3 months")
After these two data points I stopped reading. Something doesn't feel right.
sakjur 1 days ago [-]
Each file should be represented by at least three API requests, one to upload the file, one or more to check whether it's done, and one to download it. Together with the other calls of their API, I'd guess a files-to-API request ratio in the order of 1:10 rather than 1:1.
throwaway83772 1 days ago [-]
Looking at their website they support file uploads and API. They probably count those separately.
The numbers make sense to me, because I assume an API is used for high volume processing by other businesses and manual upload of files is done by individuals
They may have been inaccurate in their description, but that's what happens when regular people write articles and not scientists, engineers or LLM
Athanase000 14 hours ago [-]
It's pretty obvious from the wording and the context that it's not 550 million unique visitors.
API request is not the same thing as conversion event. You typically hit several API endpoints to carry out a single workflow (listing, authentication, upload, processing, etc).
teiferer 6 hours ago [-]
> It's pretty obvious from the wording and the context that it's not 550 million unique visitors.
The precise wording, again:
> Over 550 million people have used Zamzar since launch
How is "over X million people have done ..." making it obvious that they were not unique?
If you go to a Taylor Swift concert and the newspaper next day headlines "over 60000 people were in the audience", does that mean to you that it was obviously just a handful and they were counted multiple times?
MiscIdeaMaker99 1 days ago [-]
Who cares?
txtlog 1 days ago [-]
Maybe because it smells like self promotion? Anyway, good work if it's actually true
stevoski 19 hours ago [-]
It’s a website for a commercial product. Of course it is self-promotion! How could it possibly not be so?
teiferer 1 days ago [-]
You don't care whether an article that ends up on HN front page contains legit information?
ov3r45 1 days ago [-]
[flagged]
Rendered at 10:22:13 GMT+0000 (Coordinated Universal Time) with Vercel.
Would you consider a pay-as-you go option, perhaps with a deposit?
Right now, I won't subscribe for $12 per month (or $25 for API). But I have some use for both the interactive and API conversions and am happy to pay per conversion, so a pre-paid version (like OpenRouter) would be great.
Money is part of how you get there, but it is not a goal in itself. Many start businesses where money is the main goal, and where the amount that you are satisfied with inevitably shifts once you get richer
They also have a paid API.
"Side" in the sense that my main income coming from somewhere else, aka my "normal" job
Money that doesn’t need to be spent is freedom of time.
?
https://www.zamzar.com/converters/audio/mp3/
Great job and thanks!
[edit: what’s with the downvotes? I love the story and am curious if Claude etc have affected their business.]
They are regular users, not developers.
ok.
Rather than using this as your world-view, I think you should take this as a signal to expand the current bubble you live in. I'm guessing most of these people you refer to as "non technical" are still white-collar workers or similar?
Most people in the world, even with all the marketing towards "non developer business users", has never heard of "Claude", "Gemini", "Grok" or anything else LLM-adjacent like that, but will have some light understanding of "ChatGPT", and might actually be using some AI service via their phone or whatever, but still call that "ChatGPT" as that's essentially synonymous with "AI" in common parlance.
This is a case study of how a software company can still make money in the age of ai.
My guess is that the impact of ai on the business won't be too much as they have a good brand and a loyal customer base whereas the vibecoded ones will have a hard time getting noticed by customers. They will also have a competitive advantage by being able to handle so many different types of files, by being a one-stop-shop if you like. They will be able to attract more customers by handling really obscure file formats and in some cases by being able to do a better job of the conversion (desktop publishing and vector formats are an example). There is probably a power advantage of being able to convert n filetypes to m different formats where n and m are very large.
If they've been in business this long, they must have, by now, internal filetypes for specific classes of formats. For example, all rgb files first convert to an RGB format, and then to the destination format. Or, all transcription first convert to an internal format, and then to the destination format.
Not sure how you'd handle .ps or .pdf files in this way, though.
Tried doing it with pandoc and the result was similar but less good. (I like it when the result is more good.)
It's amazing how many formats they support. It's kind of strange that this isn't built into the computer, though. Without Zamzar, the computer still lives in the pre-Babelfish era...
My first thought was "imagine the astonishing amount of private documents and PII you could collect and store if you ran something like this for free as a honeypot". People just blithely uploading any document from their personal data to convert to another format.
And if I would have to guess it's less technical users who aren't going to try something like opening a document in the latest Libreoffice (or gimp, etc) and then exporting it to another format, using local software on their computer, first.
Why don’t you create an alternative?
> Over 550 million people have used Zamzar since launch,
There are 8.3 billion people on earth. You are saying that about 1 in 42 have used your service? I had never heard of it and I don't know anybody who had. And that's a sample that should be biased in your favor. Sorry, does not pass the smell test.
Or maybe they were not unique visitors, containing duplicates? How many? This makes me skeptical of the remaining statistics on the page.
> converting over 766 million files. We now process over 1 million API requests per day.
1 million per day now and 766 million over 20 years? At the current rate you'd achieve that in about 2 years. This would indicate a rapid growth in the last couple years or even just months. It is unclear why that would be reasonable. Especially since you'd probably have emphasized this explosion more if your service was essentially dormant for 19 years and just recently went through the roof. (Sth like "how we handled 100x growth of our SaaS business over 3 months")
After these two data points I stopped reading. Something doesn't feel right.
The numbers make sense to me, because I assume an API is used for high volume processing by other businesses and manual upload of files is done by individuals
They may have been inaccurate in their description, but that's what happens when regular people write articles and not scientists, engineers or LLM
The precise wording, again:
> Over 550 million people have used Zamzar since launch
How is "over X million people have done ..." making it obvious that they were not unique?
If you go to a Taylor Swift concert and the newspaper next day headlines "over 60000 people were in the audience", does that mean to you that it was obviously just a handful and they were counted multiple times?