Superhuman vs Hey vs Shortwave (2026): An Honest Take

The honest Superhuman vs Hey vs Shortwave summary: Superhuman is a speed tool, Hey is a reformation, Shortwave is an AI co-pilot; and none of them solve the problem of inbox volume itself. Each has a coherent philosophy. Each costs real money. And each leaves the underlying loop of "more emails arrive than you can finish" entirely intact.

If you've already read our broader take on the best email apps in 2026, you know the landscape. This post goes deeper on the three apps people most often agonize over; and why their philosophies, taken seriously, all stop one step short of the real fix.

The email app market crossed $2.1B in 2026 (FactMR), and AI adoption inside email tools doubled in eighteen months: 61% of enterprise email programs now use AI for at least one element of campaign creation, per Knak's 2026 Email Creation & AI report. Premium email apps are the part of that market where users pay out of pocket for relief. The question this post tries to answer honestly: what are you actually buying?

The frame I'll keep coming back to: email isn't broken because it's slow, ugly, or dumb. It's broken because it has no finish line. Speed makes the treadmill faster. Design makes it prettier. AI makes it smarter. None of those make it stop.

What Does Each App Actually Do Well?#

Each of these apps has earned its devoted audience by being genuinely good at one specific thing. Pretending otherwise would be cheap.

Superhuman: A Keyboard-First Speed Tool#

Superhuman's core value is that it's the fastest way to triage a Gmail or Outlook inbox if you live on a keyboard. Shortcuts for everything. A "command center" pattern that imitates power-user IDE workflows. Sub-100ms interactions on most actions. The original 2017 promise; "blast through email twice as fast"; is one Superhuman has consistently delivered for the people who actually take the onboarding session and learn the shortcuts.

In October 2025, Grammarly completed its $825M acquisition of Superhuman and then rebranded the parent company as Superhuman, folding Grammarly, Coda, and a new agent product called Superhuman Go into a single suite. The 2026 product retains the keyboard-first DNA but bolts on agentic features: Auto Drafts (writes replies in your voice without you asking), Auto Archive (silently disposes of marketing and cold mail), Auto Reminders, Auto Labels, Instant Reply, and Auto Summarize.

What it does well, distilled: if you are an executive, founder, or salesperson processing 200+ emails a day on Gmail or Outlook and you commit to learning the shortcut system, Superhuman will save you real hours per week. That's not marketing. The 2018 internal benchmark of 4 hours saved weekly holds up reasonably well for heavy users.

Hey: A Reformation of Email's Social Contract#

Hey, made by the 37signals team behind Basecamp, isn't trying to make email faster. It's trying to make email different. The flagship features are philosophical statements:

• The Screener; every new sender is held outside the inbox until you explicitly approve them. Once. Ever. If you say no, you literally never see another email from them again. It is the email equivalent of a bouncer.
• The Imbox; only the people and topics you've approved land here. Newsletters go to The Feed (a chronological feed view, like RSS). Receipts go to The Paper Trail. Marketing tracking pixels are blocked by default.
• Reply Later, Set Aside, Bubble Up; explicit triage states instead of stars and labels.

Hey treats the inbox as territory you own and have the right to defend. That stance; that you should not have to "manage" mail from senders you never invited in the first place; is its single best idea. No other email app commits to it that hard. The audience is small but loyal: 37signals reported over 30,000 paying customers within a few months of Hey's June 2020 launch, and the company hasn't disclosed numbers since; a tell that Hey is a healthy niche, not a mass-market product.

Shortwave: AI as a Senior Assistant for Gmail#

Shortwave is built by ex-Google engineers (several from the original Inbox team) and it shows. It only works with Gmail and Google Workspace, and inside that constraint it's the most ambitious AI integration shipping in 2026. The headline features:

• AI Search; natural language queries like "what did Sarah say about the Q3 budget" return synthesized answers, not just hits.
• Ghostwriter; a writing assistant that actually trains on your sent mail to mimic your voice, addressing the "AI replies sound like AI" problem more honestly than competitors do.
• Thread Summaries; long threads collapse to a few bullets. Genuinely useful for the 30-message reply-all chains nobody asked for.

Honestly, Ghostwriter is the most interesting AI feature in the category right now. I tested it on a sample of my own sent mail and it picked up specific tics; em dashes, my tendency to start replies with a one-word acknowledgment; that other tools miss. Shortwave's own marketing claims users reach Inbox Zero 45% faster using their structured workflow; a vendor stat to take with caution, but consistent with broader research showing AI-assisted inbox management cuts average response time by roughly 18% across studies cited in the Knak 2026 Email Creation & AI report.

What Do They Actually Cost in 2026?#

Pricing matters, and this is where the marketing pages start playing games. Here are the numbers as of April 2026, after the Grammarly–Superhuman restructure and Shortwave's plan refresh.

Source: Superhuman Plans, Hey Pricing, Shortwave Pricing, April 2026.

A subtlety the comparison sites usually miss: Superhuman's standalone $30/mo is now technically the "Starter" of the broader Superhuman Suite, which bundles Grammarly and Coda. If you don't want those, you're paying for them anyway. And Hey's $99/yr looks cheap until you need a custom domain; at which point it jumps to $999/yr for the first seat. Shortwave is the cheapest paid tier at $7/mo, but the free plan only searches the last 90 days of mail, which makes it a trial more than a viable free product.

How Do Superhuman vs Hey vs Shortwave Differ in Philosophy?#

Every premium email app sells a worldview. The pitch decks dress it up as features, but underneath each one there's a single sentence about what email is for. Reading the apps this way; as arguments about the inbox; is more useful than feature-by-feature comparison.

Superhuman's philosophy: the inbox is a queue, and the goal is throughput. Every interaction is shaved down. Every shortcut exists to remove a millisecond. The implicit promise is that if you process mail fast enough, you'll get to the bottom and have your evening back. The acquisition by Grammarly and the introduction of Auto Drafts modulates this slightly; now there's an "agent does it for you" layer too; but the underlying assumption hasn't changed: process faster, win.

Hey's philosophy: the inbox is your house, and most senders shouldn't be inside it. This is the most distinct of the three. Hey starts from the premise that the real problem isn't the speed at which you process mail, it's that strangers can put mail in front of you at all. The Screener is a fence. The Feed and Paper Trail are basement filing cabinets. Only the Imbox is the living room. Their bet is that if the wrong stuff never enters, the right stuff is manageable. It's the only one of the three that treats unconsented inbound mail as a problem to be prevented, not just processed faster.

Shortwave's philosophy: the inbox is intelligence-deficient, and AI fixes that. Where Superhuman optimizes user motion and Hey optimizes who's allowed in, Shortwave optimizes comprehension. Their bet is that most of the cost of email isn't volume; it's the cognitive overhead of figuring out what each thread is about, what it's asking, and what to say back. AI search, summaries, and Ghostwriter all attack that layer. The implicit endgame is an email client where the AI processes mail and you mostly approve.

These three positions are genuinely different, and a thoughtful person could choose any of them. But notice what none of them touch.

Where Do These Apps Actually Fall Short?#

This is the section the affiliate-revenue blogs leave out. Each of these apps has limitations that are worth taking seriously before you commit a year of $99–$360.

Superhuman. The learning curve is real; most users need the onboarding session to actually use it well, and people who don't fully internalize the shortcuts get a worse-than-Gmail experience at three times the price. The Grammarly bundle is great if you wanted Grammarly, awkward if you didn't. Most importantly: Superhuman makes you faster at processing email. It doesn't make there be less of it. Heavy users routinely report that they hit Inbox Zero faster, then the next wave arrives, and the cycle resumes.

Hey. The migration cost is the killer. Either you switch your primary address to @hey.com (telling every system, contact, and service of record about a new address) or you pay $999/year for the first seat of Hey for Work to bring a custom domain. There's also no Imbox-style feature that meaningfully scales: if you approve a sender once and they later flood you, there's no graceful "I let too many people in" undo without picking through The Screener history. And Hey's idiosyncratic UI is genuinely polarizing; a meaningful percentage of trialists bounce off the unfamiliar mental model in week one.

Shortwave. The Gmail-only restriction is the one that disqualifies most enterprise users on day one. If your company runs Microsoft 365, this isn't an option, full stop. The free tier's 90-day search horizon is a deliberate trial wall, not a real free product. And while Ghostwriter is the best voice-matching writing AI in email, reviews note that AI replies still don't fully match user tone for nuanced or political conversations; useful for first drafts, less so for finals.

There's a shared limitation behind all three: they assume the inbox itself is the right object to optimize. None of them ask whether processing 100+ emails a day is something a healthy human should be doing at all. Knowledge workers already spend 28% of their workweek on email, costing roughly $48,360 per worker per year in lost productivity. a stat from McKinsey Global Institute's 2012 The Social Economy report that has only worsened since. Speed, design, and AI all chip at that number. None of them name the actual variable: volume.

A Fourth Philosophy: Constraint Over Curation#

The three apps above all start from the assumption that the inbox is something to be curated; sorted, screened, summarized, accelerated. A fourth bet is different. It treats the inbox as something to be constrained.

Concretely: a constraint-first session shows you a fixed number of cards; a handful, not a hundred; and ends. Each card is a single email distilled by AI to its essence. You swipe left to clear, right to flag, up to reply, down to archive. When the cards are done, the session is done. There's no "more" button. The inbox doesn't ghost-grow in a background tab while you work.

That sounds small. It changes the loop. Speed-first apps like Superhuman optimize the rate at which you bail water. Curation-first apps like Hey try to drill fewer holes in the boat. AI-first apps like Shortwave hand you a smarter bucket. A constraint-first app says: the bucket is two gallons, the session is over when it's empty, and you go live your life.

We've written more about why this works at the level of decision fatigue and choice psychology, and about the difference between Inbox Zero and Inbox Finite. But the bumper-sticker version: the email industry has a bias toward selling you tools that match the inbox's appetite. The fix is a tool that doesn't.

"The shortest answer is doing."; David Allen, in Getting Things Done (Penguin, 2015 revised edition). The point Allen makes about open loops applies directly: a queue with no boundary is an open loop the brain never closes.

That's also what Sophie Leroy's attention residue research at the University of Washington showed in a different domain: incomplete tasks bleed into the next one. An inbox you can't finish is, by definition, an incomplete task; every minute of the day. A fixed-card session ends. The loop closes. The next thing gets your full attention.

Which One Should You Actually Pick?#

If you've read this far, the honest decision tree is:

• Pick Superhuman if you process 200+ emails a day on Gmail or Outlook, you live on a keyboard, and your bottleneck is speed of execution. Don't pick it for the AI; the agentic features are real but secondary.
• Pick Hey if you're philosophically aligned with the screener model and you can accept the migration tax. Hey is the right answer for people whose problem is "why is this stranger in my inbox at all" more than "why is my inbox so big."
• Pick Shortwave if you're a Gmail-only user, you want serious AI assistance (especially for search and tone-matched replies), and you don't mind the Gmail-only ceiling.
• Don't pick any of them if your real problem is that the inbox itself never ends. None of these three address that. A constraint-based tool; fewer cards, a clear stop, no infinite scroll; is a different category, and it's the one we're building.

One caveat: I run Swizero, so take the last bullet with appropriate salt. But the first three I genuinely stand behind based on the public pricing, my own testing, and the Reddit discourse which is more honest than most affiliate roundups.

Frequently Asked Questions#

Is Superhuman worth $30 a month in 2026?#

For users processing 200+ emails per day on Gmail or Outlook who learn the keyboard shortcuts, yes; independent reviews consistently show 2–4 hours per week saved. For lighter users, the price-to-value ratio is poor and Shortwave Personal at $7/mo is a more rational choice.

Why is Hey so expensive at $99/year for personal use?#

Hey's pricing reflects that it's a full email service (with hosting, address, and storage), not just a client on top of Gmail. The $99 price targets users who value Hey's screener-and-trays philosophy enough to migrate addresses.

Does Shortwave work with Outlook or Microsoft 365?#

No. As of April 2026, Shortwave only supports Gmail and Google Workspace accounts. There is no Outlook or Microsoft 365 integration on their public roadmap.

What's the main difference between Superhuman and Shortwave?#

Superhuman optimizes speed (keyboard-first, sub-100ms interactions, agentic Auto Drafts). Shortwave optimizes intelligence (AI search, summaries, voice-matched Ghostwriter). Superhuman supports Gmail and Outlook; Shortwave is Gmail-only. Superhuman has no free tier; Shortwave does.

Are any of these worth it if I only get 30 emails a day?#

Probably not. Premium email apps are designed for high-volume users; at 30 emails per day, native Gmail or Apple Mail handles the load fine, and the upside of speed or AI features is small. The case for paying changes when daily volume crosses roughly 80–100 messages.

Sources#

1. Superhuman Plans. Superhuman, 2026. Current pricing tiers.
2. HEY Pricing. 37signals, 2026. Personal and Work plan pricing.
3. Shortwave Pricing. Shortwave, 2026. Free, Personal, Pro, Business tiers.
4. Grammarly to Acquire Superhuman. Grammarly Blog, July 2025. Announcement of $825M acquisition.
5. Grammarly acquires AI email client Superhuman. TechCrunch, July 2025. Deal terms and rationale.
6. The Social Economy: Unlocking Value and Productivity Through Social Technologies. McKinsey Global Institute, 2012. 28% of workweek on email.
7. 85+ Email Creation & AI Statistics for 2026. Knak, 2026. 61% of enterprise email programs using AI.
8. Shortwave: New Pro, Business & Personal Plans. Shortwave Blog, 2026. Tier-gated AI search history.
9. Superhuman vs HEY Email (2026): Speed vs Philosophy. Toolfinder, 2026. Migration cost analysis.
10. Superhuman Email Review: Still Worth It? (6-Year Review). Nick Lafferty, 2026. Onboarding and learning curve.