Every once in a while, I come across a third-party app that either resets my expectations for a particular niche of software on iOS or creates an entirely new category altogether. I’ve had quite a few of these moments in the 17 years I’ve been writing app reviews at MacStories: Editorial, Workflow, Obsidian, Sky, and, most recently, OpenClaw come to mind. Over the past few weeks, I’ve had another such moment with Open Minis, a new chatbot app for iPhone and iPad that I can best describe as getting a preview of Siri AI’s agentic future, today.

Open Minis is an on-device agent that lets you use any frontier model for conversations, with a twist: unlike other AI wrappers, this app deeply integrates with all sorts of native Apple system frameworks using official APIs available to third-party developers. Open Minis can control Reminders and Music, which we have seen before, but also Calendar, Maps, HomeKit, HealthKit, and Files; it can even work with frameworks such as Vision for OCR, Apple NLP for natural language processing, the iOS clipboard, speech recognition and dictation, NFC, and Bluetooth. Furthermore, Open Minis comes with its own configurable workspace in the Files app, has an on-device memory system, and features browser-use capabilities via a (once again, native for iOS developers) built-in WebKit web view.

Effectively, Open Minis is what would happen if you rolled Claude Code and OpenClaw into one intuitive agentic experience designed for Apple users, with a powerful tool-calling harness that can improve and tweak itself, and with integrations specifically designed for the Apple ecosystem. It is the most impressive indie app I’ve seen in a while. If you were disappointed by the lack of power-user capabilities in Siri AI, this is a real, currently shipping example of what Siri AI could be if only it were powered by a frontier model with truly agentic functionalities.

The iOS CLI

The trick hidden up Open Minis’ sleeve is as ingenious as it is obvious: all the Apple framework integrations it supports are powered by command-line interfaces (CLIs) that bridge native Apple APIs (EventKit, MusicKit, HomeKit, etc.) to a code environment that modern LLMs can understand and operate with. To make all of this work behind the scenes, the developer of Open Minis had to create dedicated CLIs for each of these frameworks, which – and here comes the other trick – are driven by a model harness that has access to a built-in Linux terminal and is sandboxed to the app’s workspace.

Specifically, Open Minis runs a version of iSH, a popular Alpine Linux shell that operates a persistent, emulated Linux environment for ARM64 architecture in a sandboxed fashion. The shell inside Open Minis can download external packages from the web using Alpine’s apk package manager (which means that you can access popular tools such as ffmpeg and yt-dlp directly from the app), has its own Linux filesystem (exposed in the Files app as a native file provider location), and is able to call bridged Apple CLIs, as well as any kind of MCP server (via MCP code execution, another trend in the AI industry these days). Think about it this way: if Claude Code or Codex on the Mac can do anything across the entire filesystem by virtue of running in the Terminal app, Open Minis can do (almost) anything within its app sandbox thanks to a built-in terminal emulator – including running Python, using Git and SSH, and accessing standard Unix text utilities.

However, this doesn’t mean that, when using Open Minis, you’ll be looking at a terminal UI all the time: the underlying shell is simply what makes all of this possible behind the scenes. The app can display a picture-in-picture terminal preview that you can open to preview what it’s doing, but you don’t have to. The app’s main interface is your classic, chatbot-style UI that, depending on the model you’re using, will display interleaved thinking blocks, tool calls, and messages in a conversation transcript.

The other hidden power of Open Minis (which is reminiscent of OpenClaw, or, perhaps more aptly, a self-improving system such as Hermes Agent) is that it can inspect and modify its own workspace via dedicated Minis CLIs. For example, you can ask Minis – in natural language, of course – to add a new model or MCP server, and the app will bring up a native settings screen and ask for permissions to do so. If you want to store a password or API token, the app won’t do it in plain text: it’ll display a native, Keychain-based page where you can securely store your credentials. And more importantly, Minis’ own CLIs are used to recall previous conversations via the app’s memory system (based on Markdown files), create skills, and answer questions about the app itself. For instance, when I asked Minis to describe what Apple CLIs it has for iOS users, it happily provided a full guide:

Or when I asked to recall all the things we worked on during the day, Minis used one of its CLIs to read transcripts of previous chats (which, again, are simply stored in the Files app), recall memories, and present me with a useful summary of all the tasks we completed.

The same is true for adding new functionalities to Minis. Since it’s running a Linux shell with access to Python and can store assets in a persistent Files workspace, I figured I could ask Minis to install my own Apple Frames CLI for itself by giving it a link to my GitHub repo. GPT-5.6 Sol (my model of choice in the app) didn’t hesitate: it found all the dependencies for the Frames CLI, installed them, pulled the CLI’s official skill, and was good to go a minute later. And since Minis also comes with a native Photos integration, I was able to ask:

Grab my three most recent screenshots, randomize their frame colors, merge them, and save the image to my Temp Files folder.

A few seconds and Apple CLI calls later, Minis saved the resulting framed image to a folder of my choosing. The ability to work in any folder in the Files app outside of Minis’ sandbox is another official iOS API that the app is leveraging; we first saw the ability to “mount” specific folders from the Files app in Working Copy years ago, and the feature was later adopted by the likes of Ulysses, iA Writer, MWeb, and other document-based apps. In Open Minis, you can bookmark any folder from any file provider location in the Files app and tell your agent to save files to or read them from there.

This feels magical to me, but it’s merely one of the many examples of what is possible when you combine large, powerful models with the more-open-than-you-think, secure nature of iOS and iPadOS.

Working with Open Minis on iOS

All of this a lot to wrap your head around at first, and there’s no denying that Open Minis is a power-user tool designed for people who, like me, enjoy using frontier models and living at the bleeding edge of AI. What’s different in this case, though, is the fact that Open Minis – unlike other apps – embraces its Apple-first nature by fully supporting technologies that are made by Apple, released to third-party developers, and not being used by Siri AI at all at the moment.

It’s impossible to “review” an AI chatbot app since everyone’s use cases are so different, but I’m going to nevertheless describe a few scenarios that should give you an idea – and hopefully reset your expectations, too – of what can be accomplished with an AI that has deep hooks into iOS.

For starters, you can use any model you want in Open Minis, including the latest GPT-5.6 models, Claude Fable, anything from OpenRouter, and more. (I even asked Open Minis to create our own proxy for the OpenAI API, running on my Mac Studio, so I could use Sol in Fast mode; it works!) You can press the / button above the keyboard to trigger reasoning levels, compaction, memories, and skills; you can also create model groups, model defaults, and subagents from the app’s settings, then ask the model to use a specific set of subagents for a particular task, and it’ll do it.

Approach this app with this concept in mind: nearly everything about it is open and configurable by just talking to it. Short of rewriting the app’s code itself, you can ask Open Minis to tweak all sorts of things about itself and its own settings, and the agent will do it. Personally, I find this to be a more effective, elegant, and stable harness than OpenClaw for these kinds of things. Even though the app’s UI could use some polish (I don’t like the reasoning picker, and I wish I could collapse thinking steps by default), it is incredibly fun to tune Minis’ workspace to your liking by talking to a model. You can, in fact, just do things.

A few days ago, I was curious about the state of the temperature and humidity in our apartment. So I asked Minis:

Get me a concise report of the state of the temperature and humidity in my house, present it as a nice table.

A native HomeKit permission prompt and a couple minutes later, Sol in Minis parsed a raw dump of temperature and humidity readings from each sensor, reviewed each room configured in HomeKit, and presented the results. Siri AI, for comparison, presented…something:

I also figured it’d be cool to compare the Photos integration in Open Minis to Siri AI’s photo library access. In theory, they’re both using the same framework, with the same metadata for each photo, but Siri AI also has access to “personal context”, right? So, I asked:

Based on my Photos, would you say I was recently on vacation?

I will let the results speak for themselves:

How about HealthKit? Again, keep in mind that Open Minis isn’t employing any “hacks” or workarounds here; it’s simply accessing the same native Apple APIs and frameworks that all developers can use. I asked:

Find the photo of when I was on vacation recently and took a picture of a large ship. How much did I walk that night?

Minis started thinking about this request in a different way: it checked its memory first, and it realized that we recently discussed my recent vacations, so it could immediately short-circuit the process of finding out about those dates. Then, it started filtering my photos by those dates, found the original photo of the Palinuro vessel, and…yeah:

Inspecting the thinking steps in the request is also a fun exercise in seeing how a frontier model driven by an exceptional iOS harness can reason through a problem like this. The model – using the ‘Minis Computer’ (i.e., the aforementioned Linux shell) – fetched memories as Markdown files from the Files app, called the Apple Photos CLI, assembled a contact sheet with a grid of photos, picked two candidates, selected the best photo, isolated the date, called HealthKit for that date, and only then built an answer. And since all of this is based on deterministic code that was executed on-device, the response was fast and did not contain any hallucinations.

I then decided to go a step further: I asked Minis to take photo highlights from my recent vacations, put them on a map powered by Apple’s MapKit JS, and create a nice, interactive HTML artifact with photos on a map. I used Kimi K3 for this. Minis first asked me to securely store my MapKit developer token, then got to work. It collected photos and parsed metadata returned by the Photos app, and a couple of minutes later, there it was:

Notice how the HTML artifact itself is presented with WKWebView – another native Apple framework that lets Minis’ agent display webpages with the same WebKit rendering engine used by other web apps on iOS. I also combined web views with Apple NLP, the Natural Language Processing framework that’s available to Minis as another CLI. I asked the app to fetch my most recent iPad article from MacStories, scrape it as Markdown via Firecrawl, tag its different parts of speech with Apple NLP, and present it nicely. Not a problem:

Music is also a first-class citizen in Open Minis: by integrating with the Apple Media framework, agents can see what’s in your music library and find information about the albums, songs, and artists you’ve saved. Which means that if I ask Minis…

What Oasis singles and b-sides do I have in my library?

…Minis can inspect my library, cross-reference results with the web, de-duplicate matches, and present results in a way that Siri AI just can’t do right now since its capabilities are limited and the model can’t reason for longer on any specific query:

I can go on. Open Minis can access the native Bluetooth framework, scan nearby Bluetooth LE devices, and find out details about them:

It can send you real notifications when it’s done with a task, and you can even program those notifications to have a custom title and body text:

The app can also access your current location, extract locations from calendar events, and use the native Apple Weather APIs to give you detailed weather forecasts:

I think you get the idea.

By combining intelligent reasoning models with extensive integration on Apple platforms, Open Minis can deliver an agentic experience with no equal on iOS right now, which also serves as a fascinating preview of Siri AI’s potential future for power users. As a result of all this, while I still enjoy Siri AI as a basic voice assistant more capable than it was before, I’ve completely replaced its text-based chatbot functionalities with Open Minis powered by either GPT-5.6 Sol or Kimi K3.

iOS Agents Are Coming

It is ironic that, on the cusp of Siri AI launching to millions of users in English only in a few months, here we have a single indie developer who is shipping an app that puts Siri AI to shame when it comes to getting work done with LLMs on iPhone and iPad today – with any model you want, in any language or region.

Over the next few months, I expect more developers to ship products comparable to (and perhaps even exceeding) Open Minis, based on the same idea of an on-device agentic harness that combines LLMs with iOS-specific integrations. One worth keeping an eye on is Halo, an upcoming iOS agent from Tanmay Sonawane that aims to combine LLMs, iOS integrations, memory, and proactive features based on Andrew Karpathy’s pioneering work on AutoResearch in an elegant, iPhone-native experience. I’m in the early testing group for Halo, and it’s a very promising product that, unlike Minis, runs in a loop on-device (displayed via Live Activities) to gather insights about your conversations, upcoming tasks and events, and more.

As I wrote months ago, I’m surprised that none of the big AI labs are exploring this space more in depth. As Open Minis proves, now that models have grown incredibly capable at long-running tasks and tool calling, there’s a real case to be made for agents that can tap into the iOS filesystem, apps such as Photos and Calendar, and technologies like Vision OCR, location, and speech-to-text at virtually no cost. Can you imagine what a “real” Codex or ChatGPT Work for iPhone and iPad could be with a similar system to Open Minis’ and the agentic loop and harness that drive OpenAI’s models? What could Anthropic do with a Claude Cowork powered by Fable that could access the Files app, run CLIs, and sync your workspace files across devices?

With Siri AI, Apple is finally shipping a pretty good chatbot with world knowledge and personal context. Open Minis feels like the beginning of a whole new category of agentic apps for iPhone, which power users like myself can start relying on and fitting into their workflows today.

In the month that I’ve been using iOS 27 (launching in public beta today) and the new Siri AI, I’ve come to a few conclusions: Siri AI is a much better Siri and will (probably) be a massive success; the newfound chatbot capabilities of Siri will pose a serious threat to OpenAI and the vast majority of free ChatGPT users they have; and the new Siri is (again, probably) not for me at the moment. At the very least, it can’t be right now given its lack of support for third-party apps and technical issues I’ve encountered during beta testing.

It took me a while to accept the fact that Siri AI can’t be the kind of AI assistant I’d spend hours working with every day. But after testing iOS 27’s new Siri experience every day (and breaking it often in the process), I think that’s okay – and, arguably, what Apple is shooting for at the moment.

The new Siri AI is not designed for people who have grown accustomed to working with frontier models and agents such as Claude Code and Codex. The first version of Siri AI can’t be considered an “agent”, nor can it be compared to products from Anthropic, OpenAI, or Google that are targeting power users living at the bleeding edge of AI. If you go into Siri AI thinking that it’s Apple’s answer to ChatGPT Work or Claude Cowork for long-running tasks or deep research queries, you’re going to be disappointed – like I initially was.

Instead, Siri AI carries all the typical traits of a new Apple product: it’s a user-friendly experience designed for the masses, without any of the complexities that define a power-user tool, for better or worse. This is how Apple rolls: they don’t start from a high power-user ceiling and let products trickle down to consumers; they always start from the consumer level and refine, iterate, and extend until products trickle up to power users. This perspective isn’t meant to excuse the (many) current flaws of Siri AI. (More on them below.) It’s just the necessary lens for understanding how Apple is addressing their AI shortcomings in a way that doesn’t betray the company’s soul and design ethos.

There will be plenty more to say about iOS 27, the improved Shortcuts app, Liquid Glass tweaks, and other system enhancements later this year. After all, I have a review to write. Today’s column is merely an opportunity for me to reason through the obstacles I’ve faced with Siri AI and the way I’ve had to re-learn how to think about Siri.

Siri AI: The Good Stuff

You see, I believe my issue is that I’m too agent-pilled and accustomed to products like Claude Code and Codex at the moment. Desktop agents have fundamentally changed what I do as a creator who writes and talks on podcasts but also releases software products for my team and audience. There is no denying that Apple’s new Foundation models powering Siri are a far cry from Fable and Sol, or that they feel like models from two years ago. The AI frontier from the two big labs (and – we have to be honest with the numbers here – probably SpaceXAI and Meta now, too) is in a completely different league at the moment. I spent my first couple of weeks with Siri AI trying to convince myself that I could replace Codex with Siri AI, and that was the wrong approach. There is a potential future in which Siri AI can become a more proactive and capable agent, but that’s not what Apple is shipping this year.

If you, like me, are a heavy Claude or ChatGPT user but are curious about Siri AI, you need to reset your expectations and understand that Apple is going after something else entirely. For starters, the new Siri AI is a dramatically better Siri for typical Siri stuff. The new Siri understands your commands and nuanced requests better because of the LLMs behind it; it can often perform multiple requests in one go (such as setting multiple timers with one command); the new expressive (and customizable!) voices sound great, and they’re terrific at reading and summarizing notifications. Apple’s first advantage compared to other chatbots is right here: only Siri can integrate with the iPhone and iPad to this extent, so even if the Foundation models aren’t as smart as the competition, Siri can do a lot more for most people on their devices because neither OpenAI nor Anthropic have OSes they can control. Claude and ChatGPT are extremely intelligent inside their own apps; Siri AI is intelligent enough for routine tasks that can span an entire operating system.

The other key advantage of Siri AI is its system-wide persistence. In addition to voice mode, there’s now a full-blown Siri app that behaves like a traditional chatbot app, allowing you to revisit threads and pin specific ones at the top; but most importantly, you can now finally invoke the new hybrid Spotlight/Siri AI search UI from anywhere on iOS by swiping down on the Dynamic Island.

The lack of system-wide access to Spotlight on iOS bothered me for years, ever since I started reviewing iOS on an annual basis. Now, search is only a swipe away no matter the app I’m using, and the search box itself is smart enough to understand whether I’m looking for an app to launch or formulating a longer request for Siri AI.

---

It’s been challenging for me to “undo” advanced habits I’ve developed with agents like OpenClaw and assistants like Codex; I think MacStories readers who fall on the power-user end of the AI spectrum will probably face the same hurdles with Siri AI. However, I’ve been able to identify a variety of areas where Siri AI was faster, nicer, and more comfortable than ChatGPT or Codex Remote for practical, everyday questions.

I recently spent a few days in Gaeta, and as we were walking to dinner one night, we spotted this huge vessel docked in front of the restaurant. Typically, I would have taken a picture, started a new ChatGPT thread, uploaded the image, and asked a question about it. This time, I simply held down the Camera Control button (which now defaults to Siri AI mode in the Camera app) and captured a photo. A few seconds later, Siri AI told me I was looking at the Palinuro, a three-masted barquentine that serves as a training vessel for the Italian Navy – which was spot-on. An interaction that would have taken 20-30 seconds with ChatGPT was reduced to a five-second affair with Siri AI. Even nicer, I was able to ask follow-up questions from Siri AI’s floating overlay as well as find the chat in the Siri app later, where I continued asking questions about the ship’s history and usual locations.

Improved on-screen awareness and integration with Apple Maps has been another highlight. On our way to Gaeta, Google Maps had us drive through narrow mountain roads that we really didn’t like, even though they arguably saved us some time. For the return trip, I decided to use Apple Maps, but I noticed that it also recommended the same mountain roads. So without leaving the Maps app, I swiped down to invoke Siri and typed:

I wanna go back home (from Gaeta to Rome) but I don’t like this route. It goes through too many mountains. I wanna do the Autostrada as much as possible. Is there any way to maybe add some time to the trip and go slightly further down south to find another Autostrada entrance?

To which Siri responded:

To avoid the mountainous routes and stay on the Autostrada A1 longer, you can head slightly south from Gaeta toward the Cassino entrance or even further south toward Capua before heading north to Rome. This will add some time to your trip but provides a flatter, more straightforward drive on the highway.

Again, spot-on: the Cassino toll booth proved to be the way to go for a more comfortable trip.

A couple more things worth noting here:

• Siri AI didn’t mind my mix of Italian and English in the same request. Even though it’s only officially launching with English support, Siri AI is backed by a large language model – and LLMs are inherently good at multiple languages.
• Siri AI assembled a multi-step route for me, which I opened in Maps, and it was perfect.

After some hiccups in the first two developer betas of iOS 27, Siri AI has also become an efficient way to work with the Reminders app, which I started using as my primary task manager several months ago when I began working on my RemCTL CLI for it. In addition to the usual stuff that Siri could already do in Reminders before, such as creating tasks and fetching items due on the current day, Siri AI – thanks to the new and improved App Intents Apple added to Reminders – can now handle more advanced questions. For instance, I can ask Siri to get tasks belonging to two specific sections of a list in Reminders, and it’ll do it. Claude can also integrate with Reminders on iOS, but since its integration is powered by the old and limited EventKit framework, it can’t be as flexible as the exclusive App Intents integration now available for all apps in Siri AI.

I’ll have plenty more examples to share and analyze in my review of iOS and iPadOS 27 later this year. As things stand right now, my thoughts on Siri AI are converging around this idea: with system-wide access, an exclusive app ecosystem, and good enough performance, Siri AI doesn’t need to be a frontier-level model to be useful to millions of people.

Siri AI: The Bad Stuff

All of this, however, is not to say that Siri AI is perfect. Far from it.

This is beta software, of course, and Apple has time to improve the experience. But I’m afraid that some design choices and intrinsic limitations of Siri AI will remain in place for the next year, leaving us with a product that, despite a two-year delay, still feels very early compared to the likes of ChatGPT, Claude, or Gemini, even in their plain “chatbot” modes.

The biggest issue I’ve encountered is the inconsistency of it all. Very often, the new Siri AI still replies with a generic “Uh oh, something went wrong” error that doesn’t explain what the error was or why it happened. Sometimes, I ask a question via voice, and it errors out, but then a few seconds later, I type the same question in the app, and it’s fine.

Anecdotally speaking, these issues seem to occur more frequently when Siri “thinks” longer for a request, which perhaps suggests Apple needs to optimize the performance of their larger model on Private Cloud Compute. However, the fact that Apple wants to keep things simple compared to their competition isn’t doing them any favors here: in ChatGPT or Gemini, I can know when the model is reasoning for longer, but Siri AI’s interface is a black box. There are no reasoning selectors or hints in the UI at all. That is a fine, Apple-y approach when everything works well; but as soon as something breaks (as often happens with chatbots), the lack of clarity and controls stings.

The entire Siri app is, in fact, a bit of a mess. Its performance is subpar: I’ve noticed frequent stutters when scrolling threads, delays when typing (letters wouldn’t appear on-screen and autocorrect wouldn’t work), and broken ‘Sources’ popups displaying no actual sources for a query.

The app lets you rename and pin threads, which is nice, but it doesn’t offer basic controls such as the ability to select text in a Siri response or create folders for chats. Although the grid view looks good in marketing screenshots, in practice, I had to disable it immediately because I couldn’t quickly parse my threads and find what I was looking for.

Which brings me to indexing. As I noted above, Apple’s advantage compared to other chatbots will be Siri’s access to an on-device index that draws from the user’s apps and personal context. The problem is that I’ve yet to see an example of personal context “enhancing” any of Siri’s answers, and performing actions in or retrieving data from apps has been half broken for me since the very first beta of iOS 27.

I can’t speak about personal context because it has literally never worked in my examples. For apps, my understanding is that third-party ones will need to be specifically updated for iOS 27 in order to work with Siri AI – although some of them seem to work already, but others don’t? Apple requires developers to implement new features of the App Intents framework for Siri AI, and I’m optimistic that those who do will have their apps appear in Siri AI by September. Greg Pierce, the developer of Drafts, has added early support for Siri AI in his app, and I’ve been able to successfully create a note within the app from Siri, which presented a custom UI snippet for Drafts and completed the request in seconds. That was neat.

My problem is that, right now with Apple’s own apps, both the index and Siri’s ability to perform actions are very inconsistent. The assistant hallucinates abilities it doesn’t have, and it frequently just doesn’t work. In the examples below, I wanted to get notes from a specific folder, songs from a playlist, and a link from a note. Siri AI responded with an empty list of notes, said it couldn’t get songs from a playlist, and told me that my note didn’t have any links, respectively. (It did.)

Reminders integration started working for me in iOS 27 beta 3, but it still isn’t perfect. When I ask for tasks due today, Siri AI insists on presenting the same tasks twice, my only fault being that I enabled the integration between Reminders and Calendar:

At one point, I asked Siri to show me my overdue reminders. Siri AI struggled with that: it cut the list of overdue items short, apologized when I pushed back, then did it again anyways. I know that Siri AI is not based on a Gemini model but is rather a Gemini distillation, but I think that if you’ve ever used a Gemini model before, this kind of behavior with tool-calling should be awfully familiar.

Siri AI isn’t a replacement for agentic workflows with other LLMs that can perform multiple actions in one chain of events. Here, for instance, I asked Siri to get my latest screenshot from the Photos app and frame it with my Apple Frames shortcut. Not only did Siri AI not get the most recent screenshot from my library, but the assistant couldn’t perform two actions sequentially. At least, however, it told me; when I repeated a similar question but asked to copy the screenshot to the clipboard instead, Siri first asked me to confirm my action and then changed its mind and said it couldn’t do it.

Searching for content hasn’t been ideal on my end, either. When I asked Siri to find me the latest single by Phoebe Bridgers, its default text response was an EP from two years ago; when I asked via voice, the response was correct. (…Why?) I similarly haven’t been impressed with Siri’s ability to search for messages and emails. Siri AI can reliably find recent texts from the Messages app, but the moment I start going a little further back in time, the system seems to struggle. I wanted to know when we took some time off at MacStories last summer, which I know was mentioned in a MacStories Weekly email that I have in the Mail app. Siri AI could only find a generic mention of John recharging after some time off, but it couldn’t give me an exact date because it did not find the correct email. ChatGPT Work did it in one minute and gave me a full breakdown of the weeks we took off.

All of this begs the question:

Is Apple’s on-device index a solid technological foundation or not? Is this system actually going to work at scale, for users with dozens of apps installed, each donating their data to the index, perhaps spanning multiple years of content? And can the cloud models – which are not using MCP but appear to be based on a plain JSON schema for structured data – make sense of it all?

My early tests haven’t been encouraging, but we have until September to find out.

The Trickle-Up

A trickle-up strategy cuts both ways: it gives Apple an opportunity to start simple and refine over time, but it also means that you, the user, are never given any more control than what the product allows for. As a result, Siri AI can be magical when it works; when it doesn’t, it feels like a mystery.

I can’t say whether I’m optimistic that Apple will get all of this to work in time for the fall or not because they don’t have the greatest track record in AI. I’m willing to give them the benefit of the doubt since the product itself is genuinely interesting and different from any other chatbot on the market. I want Siri AI to work, and if Apple can deliver on their vision, I’m convinced that it’ll spell trouble for OpenAI and ChatGPT. (Let’s face it: it’s the product to beat here, both in and outside of court.)

Siri AI permeates iOS 27, and that’s a threat to OpenAI and the hundreds of millions of users they have using ChatGPT for free with GPT-5.5 Instant. I don’t think Apple Foundation Models are as good as GPT-5.5 Instant, but Siri AI is the more ergonomic and integrated product. When the underlying chatbot and its world knowledge capabilities are good enough, and they can also be invoked via voice, text, the Camera app, context menus, or keyboard shortcuts, why would I open a siloed chatbot when I can use Siri AI and get a comparable answer more comfortably? If you’re a casual user of ChatGPT and do not pay for it, I think Siri AI can be a very compelling product that provides similar answers and, unlike OpenAI’s chatbot, integrates with the iPhone apps you already use. I don’t think it’s a coincidence that OpenAI is now following in Anthropic’s footsteps to monetize the business and enterprise markets rather than the consumer one.

At the same time, my concern is that we may be collectively assuming that “normal people” won’t see the differences between the capabilities of ChatGPT and Siri AI. In this industry, we have a tendency to sell “normal people” short and paint them as some perplexed bunch of occasional smartphone users who somehow passively accept any technology that is thrown at them. I strongly disagree with that. Maybe those people don’t care about ChatGPT Work, Codex, or Fable (yet?), but they can tell when an AI assistant is useful and when it’s not. People will compare the utility of ChatGPT to Siri AI, and I’m concerned that Apple’s 1.0 product will feel like it’s too little, too late – especially since Siri AI will be launching in select regions and in English only.

In the AI scene, companies move fast and break things. Apple needs to move fast without breaking anything, anymore. That’s a tough challenge, but here we are.

Saron Yitbarek, writing on the WebKit blog:

In Safari Technology Preview 247, we’re introducing the Safari MCP server — a Model Context Protocol server for web developers that makes your web development and debugging workflow faster and more powerful. We know agents are increasingly integral to the coding process and the Safari MCP server gives your agent the ability to know how your code actually renders in the browser by connecting it to a Safari browser window.

Any MCP-compatible client can connect to the Safari MCP server. By connecting your agent to a Safari browser window, your agent can emulate what your users experience, giving it the information it needs to debug more autonomously, like access to the DOM, network requests, screenshots, and console output.

Importantly:

The Safari MCP server runs entirely on your local machine and makes no network calls of its own. It also does not have access to your personal information in Safari (e.g. AutoFill or other browser activity). When it captures page content, screenshots, or console logs, that data goes directly to the agent you’re running — not to Apple. What happens to that data from there depends on the agent and model you’re using. As with any agent you give access to your browser, only use ones you trust.

For the past few months, I’ve been using Google Chrome on my MacBook Pro and Mac Studio not because I like the browser (in fact, I really dislike Chrome’s text rendering and UI), but simply because it was the best option for agents. In Codex specifically, between Playwright, Chrome Dev Tools, and OpenAI’s own Chrome extension, I could kick off research tasks (such as vacation planning and booking a hotel) that involved a browser directly from my iPhone, letting Codex drive the research on my remote Mac.

Now, thanks to Safari’s new MCP server, I no longer have to use Chrome on desktop and can return to a unified browser setup across all my devices. Even better: it actually looks like Apple shipped the most ergonomic browser MCP for agents to date. The MCP server has dedicated tools for page extraction (including getting webpages as Markdown, based on WebKit’s own conversion pipeline), evaluating JavaScript, DOM interactions (clicking, scrolling, resizing the viewport for mobile screens, etc.), taking screenshots, and more. I set this up immediately in Codex, and I also asked Codex about comparing Safari’s MCP to its own Chrome extension and the older Playwright. The verdict: although Chrome has a richer API with Chrome Dev Tools when it comes to network requests, Codex actually preferred Apple’s leaner, more direct approach for letting an agent drive and debug a browser.

I’m really happy to see folks at Apple embrace agentic tools: between the new MCP capabilities of Xcode and now this, it looks like Apple’s software (on the Mac, of course) is becoming more and more approachable by people who are working in new ways thanks to agents. Whether you’re a web developer or tinkerer, I highly recommend checking out what Apple has released in Safari Technology Preview. More of this, please.

In the current era of coding agents becoming productivity assistants, iPadOS’ limitations are no longer defined by the lack of desktop-class multitasking or access to external peripherals. A new class of iPadOS shortcomings looms large on the horizon: the iPad’s app sandboxing and the absence of an open filesystem have relegated it to acting as a remote control for agents.

I’ve been ruminating on this thought for some time. I, of course, am very satisfied with the improvements Apple brought to the iPadOS platform throughout the 26 cycle: between new windowing, local capture, a menu bar, and the improved Files app, I can get most of my old-school work done on the iPad Pro without begrudging the experience like I used to a few years ago. What I’ve noticed this year, however, is that iPadOS 26 was an update designed to address the pre-existing low-hanging fruit of quasi-desktop efficiency; by design – and we have Craig Federighi on the record saying as much – it wasn’t intended to reinvent the wheel in any meaningful way.

That would have been fine a few years ago. The problem is that the productivity wheel is literally being reinvented by other companies as we speak. And in an unsurprising turn of events, the Mac is at the forefront of that reinvention, while the iPad is being left behind once again.

I’m writing this as someone who likes to be on the bleeding edge of AI tools. Since the release of Claude Code in early 2025, I (and, realistically, millions of other nerds whose taste aligns with mine) have increasingly moved a lot of my research, automation, coding, and other tasks considered “productivity” to desktop agents. These are powered by the same underlying technology as chatbots, but they’re considerably more powerful: Claude Code and Cowork, Codex, Perplexity’s Personal Computer, and self-hosted agents like OpenClaw and Hermes can understand intent and perform long-running tasks on your computer. They can write code, stay on a task for hours (or even days) at a time, analyze the contents of your screen, and even use a computer on their own. To believe that these AI tools are still a fad and chatbots are useless is misinformation at this point.

As of June 2026, I spend most of my days with two apps constantly open on my M4 Pro MacBook Pro: Codex and Notion. I’m a huge fan of what OpenAI has accomplished with Codex for Mac; their all-in-one package that combines thread-based chats with skills, plugins, connectors, computer use, memories, and system-wide filesystem access has genuinely allowed me to produce content for my audience that I wouldn’t have even dreamed of years ago. Since I started working with Codex, I’ve released a major update to Apple Frames, two CLIs (with two more coming soon), Shortcuts Playground, and a brand new MacStories Shortcuts Archive. These are actual products that are used by thousands of people and are generating real revenue for MacStories. I’ve also created an entire suite of MacStories-only internal apps (now used by our whole team) and am currently in the process of researching my iOS 27 review, creating an app for Shortcuts Playground, and brainstorming a different product altogether for MCP.

I’m doing all of this on a Mac because the true Codex experience isn’t even “desktop-only”: it’s Mac-only.

Let me give you some examples. To make the original version of Shortcuts Playground, I had Codex with Computer Use scrape my entire library on the old MacStories Shortcuts Archive, download over 300 shortcuts, and learn my coding style. Once I had a plugin, I also had Codex create over 1,000 shortcuts on its own, install them in the Shortcuts app for Mac, and debug them visually to find issues. For the new Shortcuts Playground plugin and app, I’m going even further: Codex created over 3,500 shortcuts spanning the entire library of Shortcuts actions while simultaneously coordinating with other Codex threads – some of which are running on my Mac Studio – to build an app, train a local model for Core AI, and constantly iterate on the new harness behind the app. Meanwhile, I’m also spending time in Codex organizing sources for my iOS 27 review, managing my tasks with Reminders and my own RemCTL, and performing general research on any topic I think of. The net result of all this is that I’m shipping more things, making more money for MacStories, saving money thanks to our new internal tools, and – most importantly – learning a lot and having fun.

In all of this, the iPad’s role is to be, once again, an accessory to the Mac – and a rather expensive one at that. I wrote last year about iPadOS’ third-party app problem, and I still think all of it applies today. What’s different in 2026 is that these powerful agentic tools that are redefining coding and knowledge work, and which are centralizing productivity inside so-called “super apps”, aren’t coming to the iPad, but not for political or financial reasons: they’re Mac-only because only macOS is mature and open enough to support them.

In a funny twist of fate, Apple spent years trying to make iPadOS catch up to macOS on the surface, only for macOS’ underlying nature to unlock a new generation of apps that are rewriting the rules of modern software. No amount of windowing and menu bars can change the fact that neither Claude Cowork nor Codex will ever be possible on the current version of iPadOS. From an open filesystem that lets agents poke around anywhere to powerful Accessibility and Screen Recording APIs for developers to features like system-wide hotkeys, dictation, and true background processes, Apple’s oldest operating system is the one that feels the freshest (and the most powerful) in the AI industry today.

In fact, the Mac is so powerful and well-equipped for this new wave of desktop agents, it even yields an amazing experience when used in a “headless” setup with no display attached to it. This may have started as a trend because of OpenClaw, but six months later, I believe that a Mac mini with Codex running on it is the best bang for your buck if you’re building an always-on agentic setup based on a frontier model that can do just about anything you throw at it.

Which brings me to the iPad’s only silver lining in this space right now.

Most of my Codex tasks are running on my Mac Studio server, but I also use Codex on my MacBook Pro. Thanks to Codex’s support for remote connections, I can chat with threads running on the Mac Studio from the Codex app on a different machine, which is a fantastic feature. Alternatively, I can use Apple’s excellent (and, again, Mac-only) Screen Sharing app to log into my headless Mac and interact with macOS with superior performance and image quality to VNC. (I also use this HDMI 2.1 dummy plug to unlock 4K resolution when headless.) Coordinating Codex threads across different projects and machines feels like the sort of thing that shouldn’t be possible, yet here we are.

If the iPad is still relevant to me in this specific new way of working, it’s entirely because of the ChatGPT app and the fact that OpenAI cared enough to design a good remote experience for iPhone and iPad users. The Codex section inside ChatGPT is a solid product and getting better every week; it’s also the only app of this kind that lets you connect to multiple computers at once and manage threads across all of them. With Codex remote on iPad, I can continue my projects from the MacBook Pro or Mac Studio, pin specific threads, and keep an eye on all my agents working in the background. In general, it’s also a way to bridge functionalities that are missing from iPadOS. Without Codex remote, I wouldn’t be able to ask my agent to reschedule my reminders and turn some emails into tasks in the same workflow; I wouldn’t be able to tell my agent to go to the Apple Newsroom, find me a specific press release, extract all images from it, and upload them to our CDN.

Perhaps Siri AI will unlock similar capabilities (although I doubt to the degree of a frontier desktop agent), but it’ll still be limited to apps made by companies that decide to support App Intents. When I’m not using my Mac and I think of something I can’t do on the iPad these days, I instantly reach for Codex inside ChatGPT. I suspect this trend will only grow as more AI companies release similar remote experiences for iOS and iPadOS.

Here’s the issue, though: these are remote companion experiences. It’s great that they exist and that OpenAI cared enough to make them great, but they’re still no substitutes for the real Codex experience on a Mac. It’s ironic if you think about it, isn’t it? A headless Mac is still more flexible and powerful than an iPad hamstrung by the closed nature of its own OS and software gatekeeping.

If you’re among the five million people using Codex on a weekly basis and will be planning your next laptop purchase in a few months, perhaps these considerations will change the calculus for you. Especially in the face of price increases and RAM uncertainty, would you be better served by spending $1,499 on a hybrid laptop that gets work done the old way with 256 GB of storage, or $1,499 on a computer where AI is a first-class citizen with 24 GB of RAM and double the storage? If we accept that power users’ priorities and workflows will be dramatically changing in the near future (if they haven’t already!), all prices being equal, which OS is best positioned for the future?

Of course, I like using both an iPad Pro and a Mac for different tasks. But let’s face it; in this day and age – and especially after last week – not everybody has the budget for two computers that have some overlap between them. Sure, I could argue that an iPad Pro + a headless Mac mini is an insane combo to get the best of both worlds these days: the nimble and delightful nature of iPadOS and the freedom to experiment with AI on a Mac. You can install apps like the latest Jump Desktop beta or the great new screen sharing app Mirage on the iPad and use your Mac mini almost as if it were connected to the iPad’s display. Set up Tailscale on your network, and the Mac’s power can be with you at all times. But that’s a lot of money, and a lot of setup, to achieve an approximation of what a single MacBook Air can handle with more memory, more storage, and a couple of Mac apps.

I’m sure that Apple is aware of the changing nature of this landscape, but in a similar vein to chatbots and Siri, perhaps they were unprepared for the tide of agents to accelerate so quickly. If the idea of putting macOS on an iPad Pro continues to be a non-starter at Apple, and assuming agent-based coding and knowledge work are here to stay, what should the company do to make sure the iPad doesn’t get left behind again? Is a sandboxed Terminal for iPad backed by a dedicated “Agent Framework” the solution? Is working with other companies to build more “remote companion” apps the way forward?

Or, perhaps more obviously: how long until Siri AI can also become a capable agent, connected to App Intents, on par with competing frontier agents?

Until, and if, that happens, the iPad will once again play the role of sidekick to macOS.

When it comes to the bleeding edge of AI and pushing the frontier of agents forward, there’s nothing like the Mac.

Update: An earlier version of this article referred to this feature as having been introduced in iPadOS 27, but it was actually introduced in the iPadOS 26 cycle. I missed it. My apologies.

---

In iPadOS’ Shortcuts app, the existing ‘Open App’ action was recently updated with the ability to launch an app with a specific window placement parameter. This means you can now automate window positions on iPad by opening a bunch of apps and programmatically selecting where their windows be placed.

To find this feature, use the ‘Open App’ action in Shortcuts and expand it to reveal the new ‘Window Location & Size’ field. Here, you can click the default ‘Full Screen’ value to reveal a host of options:

• Full Screen
• Left
• Right
• Top Bottom
• Top Leading
• Top Trailing
• Bottom Leading
• Bottom Trailing
• Left Third
• Middle Third
• Right Third

You can play around with these options to open any app and resize any window by choosing from the same presets that are otherwise shown by iPadOS when right-clicking the traffic light controls:

That was a neat discovery, but as I was playing around with the action, I realized that its practical utility would be impacted by the fact that you have to manually pre-select the app you want to open as well as its window layout. Wouldn’t it be better if you could turn this Shortcuts action into a reusable technique to programmatically get whichever app is in the foreground with no manual selection upfront, then also give it a pre-selected window preset? As it turns out, you absolutely can.

This is why I created MultiSwitcher, a simple proof-of-concept shortcut for iPadOS users that lets you resize any foreground window to a preset of your choice – no manual input necessary. The idea is fairly straightforward: you can use the ‘Get Current App’ action to get the name of the foreground app, pass it to the ‘Open App’ action, and also pre-configure that action to accept a layout variable.

MultiSwitcher is best used as a shortcut added to your iPad’s dock, so you can click it at any time and instantly resize the foreground window. For MultiSwitcher, I designed a visual menu (with some ASCII characters) that lets you preview window layouts, pick one, and resize the currently frontmost window. It doesn’t have to be this way, though: if you’d rather add multiple variations of MultiSwitcher to your dock, each with a specific window size already selected, you can also do that and skip the menu. The choice is up to you.

While these newfound window-control capabilities for iPadOS are still a far cry from the Mac’s breadth of window managers and multitasking tweaks, I find this implementation in Shortcuts to be quite elegant and intuitive for most users to grasp with no need for a third-party app.

You can download MultiSwitcher for iPadOS below.