There's a specific kind of frustration that builds when you've built something beloved, watched it get absorbed into a giant, and then spent a few years inside that giant knowing exactly what it's capable of — and what it won't do. That's the subtext behind Acme Weather, the new iPhone app from the team that created Dark Sky, launched today after years inside Apple.

Co-founder Adam Grossman is diplomatic about it in the app's introductory blog post. "We enjoyed our time at Apple," he writes. "So why did we leave to start another weather company?" The answer he gives is simple: they looked at every weather app on the market, felt unsatisfied, heard the same from friends and family, and missed the days of being a small, scrappy shop. What's between those lines is the real story. Apple is a trillion-dollar company. When you have a billion users, as Grossman told TechCrunch, "mistakes are costly." There are long software development cycles, a lot of stakeholders, and an institutional reluctance to try weird, experimental ideas. For a team that thrived on hyperlocal weather innovation, that environment has a ceiling.

So they left. And now, with co-founders Josh Reyes and Dan Abrutyn — also former Dark Sky veterans — they've built Acme Weather: $25 per year, iOS now, Android coming soon, bootstrapped, and built around a central idea that most weather apps are afraid to touch.

What Acme Weather Actually Does Differently

The premise is deceptively simple: weather forecasts are wrong sometimes, and hiding that fact doesn't help anyone. Every app gives you a single prediction — a number, an icon, a line on a graph — as if meteorology were deterministic. Acme Weather doesn't pretend. Its signature feature, Alternate Predictions, shows multiple forecast lines alongside the primary one. When those lines cluster tightly, the forecast is solid. When they diverge, the app is telling you conditions are volatile and you should plan accordingly. It's the kind of signal that meteorologists read from ensemble model outputs every day, brought to a consumer interface for the first time.

The underlying forecast model is the team's own — built from scratch using numerical weather prediction models, satellite data, ground station observations, and radar. Grossman says it's better than what they had at Dark Sky, which set a high bar. Dark Sky's minute-level rain alerts were the gold standard for hyperlocal nowcasting before Apple acquired the company in March 2020. Apple eventually shut Dark Sky down entirely in January 2023, integrating pieces of its technology into Apple Weather and launching WeatherKit as a developer API. What didn't carry over cleanly was the soul of the original product — the obsessive focus on telling you exactly when the rain will start on your specific block.

Acme Weather revives that with a few additions. Community reporting lets users submit real-time conditions using icons or emojis, filling in the gaps that radar misses — light rain, patchy snow, sudden shifts in temperature. It's the Waze approach applied to weather, and it's particularly useful during fast-moving storms when model outputs lag behind what's actually happening on the ground. Maps are deeply integrated into the core forecast view rather than buried in a separate tab, and notifications cover everything from government severe weather alerts to, via the experimental Acme Labs section, rainbow sightings and beautiful sunset predictions.

That last part matters more than it sounds. It signals what kind of company this is going to be — one that's allowed to have fun, to experiment, to ship something a little unusual and see what happens. That's not something Apple's Weather team can do at scale.

The Acquisition Pattern Worth Watching

Apple's acquisition of Dark Sky in 2020 followed a well-worn playbook: buy a beloved indie app, fold the talent and technology into the platform, shut down the original product, and call it improvement. Apple Weather is better today than it was six years ago — credit where it's due. But the transaction also eliminated a competitor, ended access for Android users immediately, and shuttered a developer API that third-party apps had relied on. The community that had built around Dark Sky lost something that Apple's built-in app, however improved, didn't fully replace.

Grossman and his team now find themselves in an interesting position: they built WeatherKit from inside Apple, which means the developer API that powers much of Apple's weather ecosystem on third-party apps is partly their work. They know exactly how it functions, where it's limited, and why building their own forecast stack for Acme Weather — rather than licensing WeatherKit — gives them more flexibility. "Most of our time has been spent on building our own forecast — our own data provider, in a way," Grossman told TechCrunch. "This lets us do things like build multiple forecasts, create any map we want, rather than having to rely on a third-party map provider."

That independence is the point. At $25 a year, Acme Weather makes money directly from subscribers, with no advertising, no data brokering, and no reliance on Apple's infrastructure. The privacy policy is unusually direct: no location history stored, no third-party trackers, no selling data to advertisers. For a weather app, which by definition knows where you are at all times, that's a meaningful commitment.

The app is currently US and Canada only, which is a notable limitation for a team that built Dark Sky into a globally beloved product. An Android version is in development, and the team is actively hiring Android developers. Whether a developer API follows — Dark Sky's paid API was widely used by third-party apps before Apple killed it — remains undecided, though Grossman has acknowledged the architecture could support it if demand materializes.

What This Means for Apple Weather

Apple's built-in weather app has a structural advantage that no third-party can match: it's free, it's preinstalled, and it's deeply integrated into iOS. Widgets, lock screens, Focus modes, Siri — Apple Weather touches everything. The question Acme Weather implicitly asks is whether "good enough and free" is actually good enough for users who genuinely depend on weather forecasts to make decisions.

The complaints about Apple Weather are consistent: delayed alerts, forecasts that don't match what's happening outside, hyper-local accuracy that still falls short of what Dark Sky offered at its peak. Apple has improved significantly since the acquisition, but there's a ceiling imposed by the need to serve a billion users across wildly different climates, use cases, and levels of meteorological interest. A team of scrappy engineers building specifically for users who care deeply about weather can iterate in ways that Apple's Weather team simply can't.

Grossman's team has done this before. They built Dark Sky into one of the most beloved apps on the App Store, sold it to the largest technology company in the world, helped transform its technology into a platform-level tool, and then walked away to do it again. The fact that they're doing it again — bootstrapped, independent, with no outside investors and a lean team — suggests they believe there's a real market for a weather app that takes uncertainty seriously rather than hiding it behind a confident icon and a single number.

They're probably right. If you've ever glanced at Apple Weather's 40% rain probability and wondered what that actually means in practice, Acme Weather is building the answer to that question. And given that the people building it spent the last several years inside Apple figuring out exactly why that question is so hard to answer, they're better positioned than anyone else to try.

Acme Weather is available on the App Store with a 14-day free trial and a $25/year subscription. An Android version is in development.