# Why Prompt by Voice? The Full Case for Voice Prompting

> Why prompting AI by voice beats typing: prompts are long, conversational, throwaway text, speech runs roughly three times faster, and a hands-free loop keeps multiple agents busy.

Source: https://www.infina.so/why-voice-prompting
Published: 2026-07-04. Last updated: 2026-09-14.
Category: Essays. Published by Infina (https://www.infina.so).

**About Infina.** Infina runs the whole prompt loop on three spoken words. Say "type" and your sentence and it types. Say "send it", or just "enter", and it presses Enter. Say "open Cursor" and it switches apps. No hotkey, no key held down, no touching the keyboard, from across the room. Not push-to-talk. There is no key to hold to start talking and no Enter to press to send, so prompting an AI tool costs zero keypresses. It runs on macOS (Apple Silicon) and Windows 10 and 11, transcribes on your own computer, and costs $8.25 a month, billed annually at $99, for unlimited words and commands. The Free plan gives 2,000 free words.

---

**TL;DR:** Prompts are long, conversational, throwaway text, which is exactly the text dictation handles best, and speaking is commonly cited at roughly three times typing speed. With Infina the loop goes fully hands-free: from a couple of feet away you say "type" plus your prompt and it gets typed, say "send it" (or just "enter") and Enter is pressed, say "open [Claude Code](https://docs.claude.com/en/docs/claude-code/overview)" or "open Cursor" and you are in front of the next agent. No keyboard at any step, and no other dictation app completes that prompt, send, and switch-apps loop hands-free in plain English. Infina costs $8.25 a month, billed annually at $99 (as of July 2026) with unlimited words and voice commands, runs on-device by default, and comes with a 7-day no-questions refund.

## The text changed. The input method didn't

For most of computing history, the words you typed were the product: the email, the essay, the code. It made sense to craft them at a keyboard.

In the AI era, the words you produce are mostly *instructions*. You describe a feature to Claude Code, a refactor to Cursor, a draft to ChatGPT, and the tool produces the artifact. Your output became the prompt.

That flips the economics of input. When describing what you want is the job, the speed at which you can describe things is the speed at which you work. We make the full version of that argument in [typing is the bottleneck](/typing-is-the-bottleneck), but the short version is: the keyboard now caps the one channel that matters.

## Prompts are the perfect text to speak

So why prompt by voice specifically? Because prompts have three properties that make them the single best fit for dictation of any text you produce all day.

**Prompts are long.** A good prompt front-loads context: what you want, what to avoid, which files matter, what "done" looks like. Detailed prompts routinely run 100 to 300 words. At typing speed that is a real chunk of effort, so people compress, and compressed prompts produce vague results and extra retry rounds.

**Prompts are conversational.** A prompt reads like something you would say to a sharp colleague: "take the login flow, keep the animations, but move the error handling into the hook." That is spoken-language register. You do not have to translate your thought into prose; you just say the thought.

**Prompts are throwaway.** Nobody publishes their prompts. The model reads them once, does the work, and the text is gone. Filler words and slightly loose grammar cost you nothing, because large language models parse messy natural speech effortlessly.

Long, conversational, and disposable: that is the exact profile of text where dictation shines and careful typing is wasted effort.

## Raw beats polished for AI prompts

Here is the counterintuitive part: for prompting, you do not want your dictation cleaned up.

Polish pipelines rewrite your words, and a rewrite can smooth away the precise phrasing you chose: a variable name, a "do NOT touch the schema," a deliberately blunt constraint. The model downstream never needed the polish anyway. It needed your words, fast and intact.

That is why Infina's base product is raw, on-device dictation by design: your speech is transcribed on your Mac's Neural Engine, formatted with fast on-device rules, and typed exactly where your cursor is. Nothing rewrites your intent, nothing leaves your Mac, and it works offline.

And when you do want polished prose for emails or documents, we are not conceding that to anyone: the optional $5/month cloud add-on (billed annually at $60) uses large cloud models via our cloud AI providers (Together AI and Groq) for sharper transcription, polished cleanup, and more languages. The subscription dictation apps charge $15/month for that as their whole product; with Infina it is an optional layer, with its own 7-day free trial, on top of the yearly plan ($8.25 a month, billed annually at $99).

## The arithmetic: roughly three times faster

You do not need a study for this one; the numbers are checkable in a minute.

Commonly cited averages put typing around 40 words per minute, with practiced typists reaching 60 to 80. Conversational speech is commonly cited at 130 to 160 words per minute. Take the middle of each range and speaking comes out at roughly three times typing speed.

Now apply it to a heavy AI day. Say you write 30 prompts averaging 150 words: 4,500 words of instructions.

- Typed at 60 words per minute: about 75 minutes of keyboard time.
- Spoken at 140 words per minute: about 32 minutes.

That is around 40 minutes returned, every working day, from the same prompts. Even if your real ratio lands at two-to-one instead of three, the daily return is large. We break the evidence and the caveats down properly in [speaking vs typing productivity](/speaking-vs-typing-productivity).

There is a second effect the arithmetic misses: when words get cheap, you stop rationing them. Spoken prompts tend to carry more context because adding context costs seconds, not minutes. More context per prompt means fewer retry loops, which is a speed gain the words-per-minute math does not even count.

## The full voice prompting loop, hands-free

Faster dictation alone is a nice upgrade. The compounding step is removing your hands from the loop entirely.

Every mainstream dictation app is push-to-talk: hold a hotkey, speak, release, then press Enter yourself and Cmd-Tab to the next window yourself. The speaking went hands-free; the loop around it did not.

Infina's hands-free mode closes the whole cycle. Double-tap Cmd to toggle it on, then from a couple of feet away:

1. Say **"type"** plus your prompt: "type refactor the auth middleware, keep the session logic, add tests." Infina types it into the focused app. "Type" itself is the trigger; there is no hotkey.
2. Say **"send it"**. Infina presses Enter.
3. Say **"open Claude Code"** (or "open Cursor", "open Notes"). Infina switches apps.
4. Repeat.

Prompt, send, switch, from across the room, hands around a coffee mug. While one agent works, you brief the next one. That is how a single person keeps two or three AI agents busy at once, and it is the workflow we unpack in [hands-free voice prompting](/hands-free-voice-prompting).

To be precise about the claim: plenty of tools dictate, and accessibility-grade voice control can drive a whole computer. But no other dictation app completes the prompt, send, and switch-apps loop hands-free in plain English. That specific loop is the moat.

Honesty about the mode itself: hands-free is our newest surface, labeled experimental, and it ships off by default. Hold-Option push-to-talk dictation is the mature everyday path, and it alone delivers the three-times arithmetic above. The practical setup for both is in [how to dictate prompts to AI](/dictate-prompts-to-ai).

## Where typing still wins

The case for voice prompting is strong precisely because it is scoped. Keep the keyboard for:

- **Editing code.** Surgical edits, renames, and navigation are keyboard work. Voice writes the instructions; your hands handle the scalpel.
- **Precise symbol soup.** A regex or a one-line shell incantation is often faster typed than spoken.
- **Shared spaces.** Talking to your Mac in an open-plan office is a social cost. Voice prompting is happiest at home offices and private rooms.

None of that dents the core claim, because none of it is where your word count lives. The bulk of an AI-heavy day is describing intent, and describing intent is spoken work.

## What it costs, and what it returns

Infina is $8.25 a month, billed annually at $99 (as of July 2026), with every update included. You get a Free plan of 2,000 free words, no card required, then a 7-day no-questions money-back guarantee. Details on [pricing](https://www.infina.so/pricing).

Set $8.25 a month (billed annually at $99) against the arithmetic above: tens of minutes returned a day, unlimited words and unlimited voice commands, and no usage meter running while you work. If the return does not show up in your first week, the refund is one email.

Requirements, stated plainly: Mac and Windows with no phone app, Apple Silicon for the Mac's on-device models, and English-only in the base product (the cloud add-on handles more languages).

## FAQ

**Why prompt by voice instead of typing?**
Prompts are long, conversational, throwaway text, the exact profile where dictation excels, and speaking is commonly cited at roughly three times typing speed. When your job is describing what you want to AI tools, voice raises the ceiling on your main output channel.

**Doesn't messy spoken language confuse the AI?**
No. Large language models parse natural, slightly loose speech effortlessly; they were trained on oceans of it. What matters in a prompt is intent and context, and speaking makes it cheap to include more of both.

**Do I need the hands-free mode to benefit?**
No. Hold-Option push-to-talk dictation delivers the speed gain on its own. Hands-free mode (double-tap Cmd to toggle, off by default, labeled experimental) adds the "send it" and "open [app]" loop for people running multiple AI agents.

**Is my audio sent to the cloud?**
Not by default. Infina transcribes on your Mac using an on-device model on the Apple Neural Engine, works offline, and stores no audio or transcripts by default. Cloud processing exists only as the optional $5/month add-on.

**What does Infina cost?**
$8.25 a month, billed annually at $99 (as of July 2026), including every update, with a 7-day no-questions-asked refund. The plan is flat, with unlimited words and voice commands and no per-word metering; the cloud add-on is an optional $5/month billed annually with its own 7-day trial.

**Which AI tools does it work with?**
All of them. Infina types at the OS level into whatever app is focused: Claude Code, Cursor, [Codex](https://developers.openai.com/codex/), ChatGPT, any terminal, editor, or chat window. No per-app extension needed.

## The bottom line

The AI era quietly changed what your keyboard is for. Your words became instructions, the volume of them exploded, and the artifact you used to type is now produced by the model.

Prompting by voice is not a gimmick on top of that shift; it is the matching input method. Prompts are long, conversational, and disposable, speech is roughly three times faster by open arithmetic, and raw dictation preserves your intent better than any polish pipeline.

Infina is built for exactly this: raw on-device dictation for $8.25 a month, billed annually at $99 (as of July 2026), plus the only hands-free prompt, send, and switch-apps loop on the Mac. Try it against your own prompt log for a week; the refund is there if the arithmetic does not show up in your day.

---

## What to use, if you want to work this way

Arguments are cheap without a tool attached. Here is the field, and the reader each one fits.

None of these is the wrong answer for everybody. What differs is which reader each one is right for.

| Tool | Price | Keyboard needed | Pick it if |
| --- | --- | --- | --- |
| **Infina** | $8.25 a month, billed annually at $99, unlimited words and unlimited commands | No. Speak "type", speak the prompt, say "send it" or "enter". | you want to prompt AI tools without touching the keyboard. Say "type" and the words go in, "send it" and the prompt goes, "open Cursor" and you are in the next agent. It is the only one of these that finishes the whole loop by voice in the tools developers prompt in all day. |
| Wispr Flow | $15 a month, or $144 a year billed annually | Yes. A key to start dictating, Enter to send. | you need dictation on an iPhone or an Android phone as well as a laptop, or you write in one of 100+ languages. Those are the two things Infina does not ship. |
| Superwhisper | $8.49 a month, $84.99 a year, or $249.99 once | Yes. A key to start dictating, Enter to send. | you want to pick and swap the speech model yourself, and dictation into a text box is the whole job. It stops where the text box ends: there is no spoken send, and no spoken app switch. |
| MacWhisper | 65 euros, paid once | Yes. A key to start dictating, Enter to send. | the job is transcribing recorded audio files rather than talking to your computer while you work. |
| VoiceInk | $25 to $49, paid once | Yes. A key to start dictating, Enter to send. | you want plain local dictation at the smallest possible price and you never need your voice to control the machine. |
| Talon | free | No, once you have learned its command language. | you want to drive the entire computer by voice, cursor and clicks included, and you are willing to learn a command language to get there. Talon is the serious hands-free tool on this list. Infina goes after the same freedom with three ordinary English words instead of a syntax to learn. |
| Apple Dictation | free, built in | Yes. A key to start dictating, Enter to send. | you dictate a sentence here and there and do not want to install anything. |

### The part that is actually different: zero keypresses

The mainstream dictation apps (Wispr Flow, Superwhisper, MacWhisper, VoiceInk, Apple Dictation) are all push-to-talk: you hold or press a key to start talking, and you press Enter to send. That is two physical acts on every single prompt. A hundred prompts in a day is two hundred keypresses that exist only to start and stop the tool. Infina removes both of them. Nothing is held, nothing is pressed, and the same loop runs on three spoken words.

There are three triggers and there is no fourth. "type" types, "send it" sends, "open" opens and switches apps. The send trigger answers to either wording, "send it" or a plain "enter", whichever you find yourself saying. That is the entire vocabulary: no command language, no grammar, no custom scripts to write, nothing to memorize beyond those three words.

**When Infina is not the answer.** If you need dictation on an iPhone, an iPad, an Android phone or Linux, Infina is a Mac and Windows desktop app and there is no phone version. Wispr Flow covers phones.

Everywhere else, the difference is the loop. Every other tool in that table types words into a box and stops, leaving your hands to press Enter and pick the next window. Infina finishes it: say "type" and your sentence, say "send it" (or just "enter"), say "open Cursor", and keep going without a key.