Follow ZDNET: Add us as a preferred source on Google.
ZDNET’s key takeaways
- Gmail Flows can now leverage AI to handle incoming messages.
- Google Workspace subscribers and paying AI customers gain access to these new capabilities.
- The monthly processing caps could severely restrict those with high email volumes.
Google gives, and Google takes away. I recently discovered a clever new AI-powered feature in Gmail that could be a game-changer for people drowning in email. That’s the giving part. But the feature comes with tight restrictions on how many emails it will actually process before shutting down. That’s the taking away part.
Last December, Google unveiled Google Workspace Studio, a tool designed to automate a range of tasks across the platform. At the time, you needed a Workspace account to access it. That’s no longer the case. Studio is now available to a broader audience.
Workspace Studio for nearly everyone
When I wrote about Gemini in Gmail back in April, Workspace Studio Flows (Google’s term for these small automation scripts) weren’t accessible on my $20/month Google AI Pro plan. But last week, I spotted a new icon at the top of my Gmail screen.
Gemini explained that this addition is tied to a major side-panel update rolled out in May, which made its way to so-called “premium” user accounts in late May and June. Premium here means you’re paying for either the $20/month Google AI Pro tier or the $100/month Google AI Ultra tier.
I subscribe to Google AI Pro, so I have the magic icon. My wife doesn’t pay for that plan, so her Gmail doesn’t include the feature I’m about to walk through.
Also: How to use AI to accurately troubleshoot your PC
As for whether the paid AI plans are worth the money, you do get expanded Gemini access with the premium tiers. My wife uses the free version of Gemini frequently. She mostly sticks to text-based tasks. When I tried generating more than a handful of Nano Banana images, I immediately hit usage limits. That’s what originally pushed me to upgrade to Pro.
All things considered, I’d say the new Workspace Studio Flows feature absolutely justifies the cost of AI Pro — except for those limitations. That caveat undercuts the value, particularly for power email users (the very people who stand to benefit the most).
The side-panel interface
Once you click the icon, a Flow side panel appears. While Google refers to the scripting tool as Workspace Studio, the scripts it produces are called “Flows.”
Google supplies starter scripts you can use as templates for building your own flows.
To unlock the full power, click “Do more in Studio,” where you can do (surprise) more.
Do more in Studio
This is where things get genuinely interesting. When you open Studio, you’re greeted with additional sample options to help you get going.
Selecting the “Get news headlines summarized daily” flow brings up the Studio scripting interface. For this first example, the flow sends notifications through Chat rather than Gmail.
This highlights an important detail: Studio flows aren’t limited to Gmail alone. They work across Chat, Docs, Meet, and other services. I’m focusing on Google’s Gmail implementation because it represents a massive improvement to a filtering system that had barely evolved in over 20 years.
As far back as 2023, I wrote about wanting Gmail to harness AI more effectively for qualifying and sorting incoming messages. After 31 long months, the system finally does. Well, aside from that “taketh away” caveat we’ll get to shortly. Grumble.
Define a trigger
Every Flow needs to start with a trigger, which Google calls a starter. My goal is to have Google automatically tag all press releases and promotional emails that have genuine journalistic value. I’ve built countless filters over the years. But since every sender pitching me stories for ZDNET’s audience uses a completely different email format, no rule-based tool could ever fully solve the problem.
Also: Claude Fable 5 secretly throttled AI researchers, and the internet went wild
But AI can come close. To do so, though, AI needs to examine every single email that lands in my inbox. So, I set my starter trigger to “When I get an email.” Unfortunately, this tool can’t be applied to existing emails — but hey, I’ll take what I can get.
You can configure the tool to scan all incoming emails or only those matching standard filter criteria, such as particular keywords or text patterns. Since the latter approach has never been thorough enough for my needs, I opted to scan every message. Then I want Gemini to apply its considerable intelligence to decide whether each email represents a legitimate story pitch.
Choose an action
Once your Flow kicks off with a trigger (in our case, a new email), it’s time to define what it should do. My initial instinct was to use the Ask Gemini action.
But once you run
To get that working, you’ll need to set up a sequence of tests and sub-actions. As I scrolled further, I spotted the “Add labels with Gemini” option, which seemed much more closely aligned with what I was trying to achieve.
To activate this feature, you’ll need to enable AI-powered labels. Studio requires you to keep the six pre-existing labels in place, though you’re not obligated to switch them on. If you’d like to craft your own AI-driven label, just click the button.
From there, you can define the text string for your label along with the underlying prompt. I enlisted Gemini to help me draft a prompt. After a handful of attempts, I landed on something I was happy with.
Once your Flow is built, you can put it to the test. First, you’ll need to select an email to run the test against. Studio will suggest a few recent messages, but you can also type in a search term to surface other emails to use as test data. Pick one, then run the test.
And that’s the whole process. In theory, you could build AI-powered filters to handle just about any email workflow challenge you can think of.
That said, there are some constraints. For instance, Flows can generate draft responses, but you’ll still need to manually click to send them. There’s actually some wisdom in that approach — if your prompt is too broad, you have no idea what the AI might fire off to the people emailing you.
All in all, I think this is a genuinely solid upgrade to Gmail. Except…
The Google giveth, and the Google taketh away
The feature I’ve outlined here doesn’t come without caps. Like every other cloud-based AI offering, there’s a finite number of times you can tap into the AI before the provider pulls the plug.
According to Google’s usage limits page, the $20/month Pro tier includes 2,000 flow executions per month. Put another way, your script can process 2,000 incoming emails each month. Beyond that, the service stops working.
This month has been particularly heavy, but it isn’t out of the ordinary. Thanks to all the Fable coverage, I received 7,724 emails in a single week. Normally I get a few thousand per week. At that rate, the $20/month Pro tier would last me roughly a week.
Also: The cloud vs. clouded leopard: America’s data center backlash on display at Nashville Zoo
During a busy month — major AI announcements, Prime Day, the winter holidays, and so on — I can easily surpass 10,000 emails. That exceeds even the $100/month Google AI Ultra tier, which caps out at 10,000 flow executions per month.
Whether this feature makes sense for you likely depends on whether you fall into a sweet spot. You need enough email volume to warrant AI assistance, but not so much that you blow past the monthly execution cap.
I’d almost certainly exceed my limits even if I shelled out a hundred bucks every month — and that’s with just a single flow. Building a whole suite of flows to truly let AI help organize my inbox for maximum productivity simply isn’t feasible under Google’s current restrictions.
One quick heads-up: Gemini pointed out to me that these monthly limits are suspended until July 1. Google has been letting users run fairly wild with their flows, but that changes sharply on July 1. The same limits page referenced earlier states, “Promotional access. Limits enforcement begins on July 1, 2026.”
Also: I tried Google Drive’s new AI cleanup tool to fix 14 years of storage clutter – here’s the result
Mid-twentieth-century comedian Bob Hope is widely quoted as saying, “A bank is a place that will lend you money if you can prove that you don’t need it.” So here’s my straightforward takeaway. Flows have tremendous potential. They’re impressive even in this early form. But if you genuinely need them because you’re drowning in email, you’ll hit a wall.
Perhaps when — or if — AI costs drop, that calculus will shift. At the very least, we know the technology exists and is already deployed. If you’re a Goldilocks-level email user, this capability is a fantastic tool. For everyone else, the feature will likely become yet another source of AI-related frustration.
Would a 2,000-email monthly Flow cap be sufficient for your inbox, or would you bump up against the ceiling too fast? Share your thoughts in the comments below.
You can follow my day-to-day project updates on social media. Be sure to subscribe to my weekly update newsletter, and follow me on Twitter/X at @DavidGewirtz, on Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.
