I have a confession. At the time I started building speakeasy, I had 347 articles saved in my read-later app. I know the exact number because I counted them the night I decided something had to change.
Three hundred and forty-seven articles. Essays on technology, pieces about business strategy, newsletters I genuinely wanted to read, research I needed for my work. Saved with the best of intentions. Never opened. A growing monument to the gap between my appetite for information and the time I had to consume it.
I suspect this sounds familiar.
The Problem Nobody Talks About
The internet is producing the best long-form writing in history. Independent writers on Substack are doing work that puts traditional journalism to shame. Newsletters in every field deliver concentrated insight directly to your inbox. The quality and breadth of written content available for free, or for a few dollars a month, is extraordinary.
The bottleneck is not supply. It is consumption.
I read constantly during my workday -- emails, Slack messages, documentation, code. By the evening, when I theoretically had time for the reading I actually wanted to do, my eyes were done. I would open an article, read two paragraphs, and feel the screen fatigue settle behind my eyes like fine sand. The article would get saved. The list would grow.
Meanwhile, I was spending an hour a day commuting with nothing but the same rotation of podcasts playing. Walking the dog for 30 minutes. Cooking dinner. Doing laundry. Hours of time where my ears were free and my eyes were not.
The obvious solution was text-to-speech. Take the articles I wanted to read and convert them to audio I could listen to during those dead hours. Simple.
I tried everything on the market. Here is what I found.
Speechify was the most polished option, but the pricing stopped me cold. At the time, it was $139 per year for the features I actually needed. That is more than I pay for most of my streaming services combined. For a tool that converts text to audio. I could not justify it, and I resented being asked to.
The free alternatives had the opposite problem. The voices were bad -- robotic, flat, with that tell-tale synthetic quality that makes your brain fight the content instead of absorbing it. Listening for more than five minutes was an endurance test, not a pleasant experience.
Apple's built-in screen reader was technically functional but clearly designed for accessibility, not casual content consumption. The voice quality was adequate for reading short notifications, not for listening to a 3,000-word essay on your morning commute.
Pocket had a listen feature that was decent, but it only worked within their ecosystem. If I wanted to listen to a tweet thread, a Substack post I received via email, or a blog post from a niche site, I was out of luck.
Every tool I tried was either too expensive, too limited, or too unpleasant to use. None of them solved the actual problem: I want to paste a URL, get a natural-sounding audio version, and listen to it when I have time.
Building the First Version
So I built it myself.
The first prototype was embarrassingly simple. A URL input, a text extractor, and a call to an AI voice API. Paste a link, get audio. That was it. No library, no sync, no speed controls. Just the core conversion.
But even that crude version changed my behavior overnight. I converted five articles before bed, listened to all of them during the next morning's commute, and realized I had consumed more of my reading list in one car ride than I had in the previous two weeks.
The iteration from there was driven entirely by my own frustration. What was missing? What did I actually need?
Speed control came first. I found that 1.5x was my sweet spot for most content, with 1x for dense material and 2x for newsletters I was skimming. The ability to adjust on the fly -- slow down for a complex passage, speed up for a familiar topic -- transformed the experience.
iCloud sync came next, and it was by far the hardest technical challenge. I wanted to convert an article on my phone and have it available everywhere, automatically, without creating an account or managing credentials. iCloud Drive was the obvious answer, but the implementation was anything but straightforward. The edge cases around sync timing, conflict resolution, and offline access took weeks to get right.
The library was the third critical piece. I needed a way to organize my audio articles -- not just a flat list, but something that felt like a personal audio library. Metadata from the original articles (title, author, source, publication date) had to be preserved so that browsing the library felt natural.
RSS feeds came later, when I realized I was repeatedly converting articles from the same sources. Why manually paste URLs when I could subscribe to a feed and have new articles available as audio automatically?
Each feature was born from the same process: use the app daily, notice what is missing, build it.
The Design Philosophy
Early on, I made a decision that shaped everything: speakeasy would do one thing exceptionally well rather than many things adequately.
The text-to-speech market is full of apps that try to be everything. They add AI summarization, translation, note-taking, social features, flashcards, and a dozen other capabilities. Each addition makes the app more complex, more confusing, and more expensive. The user pays for features they never touch.
speakeasy converts URLs to audio. It does this quickly, with high-quality voices, and it organizes the results in a library that syncs across your devices. That is it.
This focus has practical benefits. The app is fast because it does not have to load a dozen feature modules. The interface is clean because there are not sixteen tabs competing for your attention. The price is fair because you are not subsidizing features you do not use.
It also means I can obsess over the details that matter. Voice quality. Extraction accuracy. Speed control responsiveness. Queue management. The small interactions you repeat hundreds of times. When you are building one thing, you can make that one thing genuinely great.
What I Have Learned
Building speakeasy has taught me something I did not expect: the format you consume content in dramatically shapes what you consume.
When I was a reader-only, I gravitated toward shorter, punchier pieces. Long-form essays felt like commitments. A 4,000-word feature article was something I had to set aside time for, which usually meant I never got to it.
As a listener, my consumption patterns inverted. Long-form content is now my preference. A 20-minute essay is perfect for a commute. A 30-minute deep dive fills a workout. The longer pieces I used to avoid are now the ones I most enjoy, because the format matches the time I have.
I also read more diversely. When reading required dedicated screen time, I was selective to the point of narrow. Now that I can listen anywhere, I follow more writers, subscribe to more newsletters, and explore more topics. My information diet is broader and richer than it has ever been.
Where We Are Going
I am not going to lay out a detailed product roadmap here. That is not how indie development works anyway -- the roadmap emerges from daily use and user feedback.
What I will say is that speakeasy's direction is guided by a simple question: what else can we make effortless? The core conversion is fast and reliable. The library works. Sync works. RSS works. The next layers are about reducing friction further -- making it easier to discover, organize, and consume the content that matters to you.
Every feature decision runs through the same filter: does this serve the person who wants to listen to their reading list, or does this serve a pitch deck? If it is the former, it goes in. If it is the latter, it does not.
An Invitation
If you have made it this far, you are probably one of the people speakeasy was built for. You have a reading list that never shrinks. You have time in your day that is acoustically available but visually occupied. You have tried other tools and found them too expensive, too limited, or too unpleasant.
Try speakeasy. It is free for 3 articles a week. No account required. Paste a URL, pick a voice, and listen.
If it changes how you consume content the way it changed things for me, you will know within the first commute. And if you have feedback -- what is working, what is not, what is missing -- I want to hear it. This is still very much a product being built by someone who uses it every day, and the best features have come from users who cared enough to tell me what they needed.
Your reading list is not going to read itself. But it can listen itself, if you let it.