Nothing Launches AI Dictation - But Can Essential Voice Beat DictaFlow on Desktop?

The smartphone world just got a new dictation contender. On April 23, 2026, London based hardware company Nothing launched Essential Voice, an AI powered speech to text feature for Nothing Phone (3) users. It strips out filler words, cleans up speech in real time, supports more than 100 languages, and lets you dictate into basically any app on your phone.

The reaction has been warm. TechCrunch, 9to5Google, and Android Central all covered it. Nothing has been building toward a voice first interface, and Essential Voice is described as one of the building blocks for whatever they're working on next.

For phone based dictation, it looks genuinely good. What it doesn't solve is the desktop problem.

What Nothing's Essential Voice actually does

Essential Voice uses Gemini 3 Flash under the hood. You tap the Essential Key on your Nothing Phone (3) or activate it from the keyboard, speak naturally, and the feature cleans up your words before they hit the screen. Filler words like "um" and "uh" disappear. The text comes out looking like you typed it on purpose. It supports real time translation across more than 100 languages and lets you save personalized text shortcuts.

It's a solid step up from standard phone dictation. Most voice to text on Android just transcribes exactly what you said, stumbles and all. Essential Voice adds the cleanup layer that makes the output actually usable.

The catch: it's on Nothing Phone (3) right now, rolling to Phone (4a) Pro this month and the regular Phone (4a) in early May. If you don't have a Nothing phone, it doesn't exist for you. And if you spend most of your day at a Mac or Windows machine, a phone dictation feature doesn't move your workflow much.

The desktop gap that phone AI can't close

Speaking is roughly three to four times faster than typing. Nothing's own announcement makes this point to justify the product. The same math applies to desktops and laptops, where most professional writing actually happens.

Emails, Slack threads, long documents, code comments, meeting notes, these aren't written on phones for most knowledge workers. The keyboard is still the default input method on desktop, and modern AI dictation tools have gotten good enough that sticking with the keyboard is mostly a habit, not a necessity.

A clean phone dictation tool doesn't change that. You still switch devices, break context, and re integrate content. The friction is elsewhere.

What desktop AI dictation actually looks like

DictaFlow works differently. You hold a hotkey, speak, release, and your words appear wherever your cursor is, Gmail, Notion, Slack, a code editor, wherever. There's no separate recording window, no clipboard, no switching to a different input mode. It runs on Mac, Windows, and iOS, so the same workflow follows you across devices.

The mechanic that separates it from most tools is hold to talk. You're in control of exactly when transcription happens. No accidental sentences, no background noise getting picked up, no pausing and wondering if the thing is still listening.

The other standout feature is mid sentence correction. While you're dictating, if you misspeak, just say your correction keyword and DictaFlow deletes back to the error and keeps recording. No stopping, no keyboard, no mouse click. It's called Actually Override, and it's the kind of detail that makes or breaks a dictation workflow once you're doing it for hours a day.

For doctors, lawyers, and anyone working in Citrix or Remote Desktop environments, DictaFlow also works there via keystroke simulation. No audio redirection setup, no IT changes. That's a significant gap versus most modern dictation tools, which fail silently in locked down environments.

How Essential Voice fits the bigger picture

Nothing's announcement is worth watching. They're explicitly describing Essential Voice as a building block for a future voice first interface. That's an interesting bet. Google has been pushing harder on voice input too, and the general direction of the industry is toward reducing typing friction everywhere.

For people shopping for a phone upgrade who care about voice input, Nothing Phone (3) with Essential Voice is now on the shortlist. The feature sounds genuinely polished.

But the upgrade cycle for phones is measured in years, and the bulk of professional output happens on desktop hardware that isn't going away. The interesting question isn't whether phone dictation will improve, it clearly will, it's whether desktop AI dictation ever gets the same attention.

So far the dedicated desktop tools are ahead. DictaFlow at $7/month versus Nothing's phone only feature, which requires actually owning the hardware, isn't a straightforward comparison, but the use case for desktop hold to talk remains strong and underserved by phone first products.

The practical question

If you're already on a Nothing phone and you write a lot of messages and short form content on your phone, Essential Voice is a free upgrade worth using. Turn it on and see how it fits your habit.

If you spend most of your writing time on a desktop or laptop, which most professionals do, a phone feature doesn't solve the bottleneck. Try DictaFlow free and see how hold to talk changes the way you work at your actual keyboard. The comparison page at dictaflow.io/comparison.html lays out how it stacks up against Wispr Flow, Dragon, and the other main options if you're evaluating what to pay for.

What's your current setup for dictation? Are you using phone voice typing, a desktop tool, or just grinding through the keyboard? I'm curious how people are actually handling this in 2026.