Short answer: Acrobat has no built-in dictation, so you use a system-wide tool. Click into the comment box first so it has a cursor, hold your hotkey, speak, and release. Voice Keyboard Pro types the text at the cursor in any Acrobat field, with no plugin required.
Open any PDF that has been through a real review cycle and read the comments. Most of them are three words long. "Unclear." "Check this." "Is this right?" A few have a question mark and nothing else. The person who left them read the paragraph carefully, formed a specific objection, understood exactly why the sentence would cause a problem downstream, and then typed "unclear" into a small yellow box.
That gap between what the reviewer thought and what the reviewer wrote is not laziness. It is an input cost. The thought took two seconds. Typing it out would have taken forty. Multiply that by sixty comments across a hundred-page document and the shortest possible comment becomes the only sustainable comment. So the reviewer writes "unclear," and two weeks later the author reads "unclear" and has no idea which part was unclear or why, and the two of them schedule a call to reconstruct a thought that already existed in complete form.
Dictation closes that gap directly. Most adults type somewhere around 40 words per minute in real conditions, with pauses to think and fix. Comfortable speech sits at 130 to 150. The comment that costs forty seconds to type costs eight seconds to say. At eight seconds, "unclear" becomes "this says the deadline is thirty days but section four says fifteen, and I think section four is the one the client agreed to."
Acrobat does not have dictation, and that is fine
Adobe Acrobat has no voice input feature of its own. There is no microphone button in the comment toolbar, no speech panel, no plugin to hunt down. This surprises people who assume a mature document tool must have one somewhere.
It does not need one. Acrobat's comment boxes, form fields, sticky notes and reply threads are ordinary text inputs. Anything that types at the system cursor types into them. That is the entire mechanism: Voice Keyboard Pro lives in the Mac menu bar, you hold a hotkey, speak, release, and the text lands wherever the cursor already is. Acrobat never knows dictation happened. It receives text.
This matters more in Acrobat than in most apps because Acrobat installations tend to be conservative. On managed machines, in legal and engineering and finance environments, the PDF stack is often locked down and audited. A workflow that requires installing an Acrobat plugin, changing the trust settings, or adding an extension to a certified environment is a workflow that will not survive the conversation with IT. A menu bar app that types characters does not touch Acrobat's installation at all.
Setup takes about a minute
- Install Voice Keyboard Pro and grant microphone and accessibility permission when macOS asks. Accessibility permission is what allows text to be inserted at the cursor in other applications.
- Pick a hotkey. In Acrobat, pick carefully. More on this below.
- Open a PDF, click into a comment box so you can see the caret blinking, hold the hotkey, say a sentence, release.
There is no Acrobat-side configuration. Nothing in Preferences, nothing in the Comment tool settings, nothing to enable per-document.
The hotkey collision to avoid
Acrobat supports single-key accelerators for tools. With that option on, pressing S can switch you to a selection tool and H to the hand tool, without any modifier. That is a fast way to work and a fast way to collide with a dictation hotkey.
The safest choice is one of the function keys in the F13 to F19 range. They exist in macOS, almost nothing binds them by default, and no application built around single-letter tool shortcuts will fight you for them. If your keyboard does not expose those keys physically, a modifier combination that includes Control and Option is the next best thing. What you want to avoid is a bare letter or a common two-key combination that Acrobat, macOS or your text expansion tool has already claimed.
If you also dictate into other Mac apps, use the same hotkey everywhere. The muscle memory is worth more than any per-app optimization.
The first thing that goes wrong: no cursor, no text
This is the single most common first-run failure in Acrobat, and it confuses people because nothing appears to be broken.
Selecting text in a PDF is not the same as placing a caret. When you highlight a sentence, Acrobat has a selection. It does not have an insertion point. If you hold your hotkey and speak at that moment, there is nowhere for the text to go, and either nothing happens or the text lands somewhere you did not intend.
The order that works is always the same:
- Make the annotation first. Add the sticky note, add the highlight, or start the reply.
- Click inside the comment box that appears, so the caret is blinking in it.
- Then hold the hotkey and speak.
Two related traps in the same family. First, Acrobat's comment popups can close when you click elsewhere in the document, so if you click into the box, then click the page to scroll, then dictate, you may have lost the field. Scroll with the trackpad or the scrollbar instead of clicking the page. Second, the Comments panel on the side and the comment popup on the page are two ways to reach the same text; use whichever one keeps a stable caret for you, but be consistent, because the two behave slightly differently when focus moves.
If text lands in the wrong place, the cause is almost always focus rather than the dictation tool. There is a broader walkthrough of that failure mode in our guide to Mac dictation not working in third-party apps.
The rule: speak the reasoning, type the references
This is the boundary that makes dictation safe in a review context, and it is worth stating precisely.
Speak everything that is prose. Your objection, your reasoning, the consequence you are worried about, the alternative you would prefer, the question you want answered. Prose defends itself. If a word comes out wrong in a sentence, the sentence reads oddly and you catch it, or the author catches it, and the meaning survives.
Type everything that is an identifier. Section numbers, clause references, page numbers, figure and table numbers, drawing revisions, part numbers, dollar amounts, dates, percentages. These have no redundancy. If "section 4.2.1" comes out as "section 4.2.11," it does not look wrong. It looks perfectly correct, and it stays wrong through every subsequent pass, and someone eventually acts on it. A wrong sentence announces itself. A wrong number does not.
In practice this is less awkward than it sounds, because the two kinds of content naturally separate. You type "Section 4.2 conflicts with 7.1 —" then hold the hotkey and speak the explanation. The identifier is four keystrokes. The explanation is the part that used to cost you forty seconds. The same principle applies to dictating numbers and dates anywhere else, and it is the reason careful dictation users end up with a permanent habit of verifying digits by eye.
Five places this actually pays off in Acrobat
1. Comment and sticky note bodies
This is the highest-value surface and the one people give up on first. The useful review comment has three beats: what is wrong, why it matters, and what you would do instead. Almost nobody writes all three by hand because three beats is a paragraph and a paragraph is a minute of typing per comment.
Spoken, it is one breath. "The termination clause here gives them thirty days but the schedule assumes we can pull out in fifteen, so if this stays as written we are on the hook for two extra weeks of vendor cost. I would match this to fifteen or change the schedule assumption."
That comment resolves itself. The author reads it once and knows what to do. Nobody schedules a call.
2. Reply threads in shared reviews
In a shared review, the replies are where the actual decision happens, and they are usually the thinnest text in the document. "Agreed." "Done." "See above." A reply that says why you agreed is worth ten times more six months later when someone asks why the clause reads the way it does, because the PDF becomes the record of the decision and not just the record of the change.
Dictating replies also changes the pace of a review. When responding properly costs eight seconds instead of a minute, you respond to all of them in one sitting instead of picking the three that seem important and leaving the rest to rot.
3. Marking up drawings, specs and technical documents
Engineering and construction reviews carry the worst version of the terse-comment problem, because the documents are dense and the comment count is high. A drawing set can absorb two hundred markups. At a minute each that is over three hours, so people write "?" and move on.
The speak-the-reasoning rule matters most here. Speak the observation about the detail, the clash, the constraint you are worried about. Type the grid reference, the drawing number, the revision, the dimension. The dimension is exactly the kind of value that reads as correct when it is wrong.
4. Filling long PDF forms
Form fields are ordinary text inputs, so dictation works in them, with one strong caveat: forms are mostly identifiers, and identifiers are the thing you should be typing.
The split is straightforward. Narrative fields, explanations, descriptions of incident or scope or purpose, the box that says "describe in your own words," all of these are dictation surfaces and often the longest fields on the form. Account numbers, policy numbers, postal codes and dates are not. If you dictate address fields at all, verify every digit by eye afterward, for the reasons laid out in our piece on dictating mailing addresses.
One field, one dictation. Do not try to speak your way across a whole form in one pass, because pressing Tab moves focus mid-sentence and you will end up with half a sentence in the wrong box.
5. Everything around the PDF
The document is rarely the whole job. There is the email that carries it, the summary of what changed, the note to the client explaining which comments are blocking and which are preferences, the internal message asking a colleague to look at page twelve.
Because the tool is system-wide, none of that requires a context switch. You finish the markup, click into Mail or Slack, hold the same hotkey, and speak the cover note while the document is still fresh in your head. That last part matters more than it sounds. The summary you write ten minutes later is measurably worse than the one you write while the reasoning is still loaded, and the reason people write it later is that at that moment they are tired of typing.
Smart Vocabulary is what makes it stick
The reason people abandon dictation in professional documents is almost never general accuracy. It is that the eleven words that matter most in their field come out wrong every single time. A transcription engine knows ordinary English extremely well and has no way to know your client's name, your firm's internal terms, the name of the standard you cite in every third comment, or the surname of the counterparty.
Voice Keyboard Pro's Smart Vocabulary is a personal dictionary with replacement rules for exactly this. You add the terms you actually use and they stop being a problem.
Build it from real corrections rather than in advance. Dictate normally for two days, note what you keep fixing, and add those. For most PDF reviewers the list is short and boring: party and client names, project or matter names, a handful of standards or regulations, product names, the surnames of the four people whose comments you reply to constantly. Twenty entries typically covers it.
You can also use replacement rules for expansions you type constantly, so a spoken short form produces the full phrase you would otherwise type out sixty times. Our overview of how a custom vocabulary learns your words covers the mechanics. If you review documents dense in initialisms, it pairs with our guide to dictating acronyms and abbreviations.
The two-pass method for long documents
Reading and writing at the same time is what makes long reviews exhausting. You read a paragraph, form an objection, switch into composing mode to type it, lose the thread, and read the next paragraph with half your attention still on the comment you just wrote.
A method that works well with dictation: read a section straight through with the highlighter only. Do not write anything. Just mark the places that bother you. Then go back to the top of the section and dictate a comment on each highlight in one pass.
The second pass goes quickly because you already know what you think, and the comments are noticeably better because you are writing with the whole section in view rather than reacting sentence by sentence. Reviewers who annotate research literature use the same shape, which we described in the context of dictating notes in Zotero.
On the iPhone and iPad
Acrobat runs on iOS and iPadOS, and a lot of PDF review happens on an iPad with a pencil in hand. Voice Keyboard Pro's iPhone keyboard has a built-in mic button, so anywhere you can bring up a keyboard, including a comment box in a PDF app, you can dictate instead of typing on glass.
This is where the input-cost argument is most extreme, because typing a paragraph on a touch keyboard is slower and more unpleasant than typing it on a Mac. The comment you would never type with your thumbs is fifteen seconds of speech.
Voice Edit helps here too. If a sentence came out wrong, you speak the change rather than performing the fine-motor task of placing a cursor between two words on a touchscreen. And the two-way translation, which covers 24 languages while you dictate, is genuinely useful when the review is cross-border and the comment needs to land in the other side's language. Treat translated output as a first draft. A contract comment that will be relied on should be read by someone fluent before it goes out.
Troubleshooting
Nothing appears when I speak. Nine times out of ten there is no caret. Click inside the comment box until you see it blinking, then dictate. If the popup keeps closing, use the Comments panel on the side instead.
The text goes into the document instead of the comment. Same cause. Focus was on the page, not the field.
My hotkey switches Acrobat tools. Single-key accelerators are on. Move your dictation hotkey into the F13 to F19 range, or turn the accelerators off in Acrobat's preferences if you do not use them.
Accuracy drops when I am on a call. A Bluetooth headset switching into its low-bandwidth microphone mode degrades the audio the engine receives. Use the Mac's built-in microphone or a wired input for dictation, and check which input device macOS is actually using. Our guide to microphones for dictation goes deeper.
Names come out wrong every time. Add them to Smart Vocabulary. This is not something to keep fixing by hand.
Long comments trail off at the end. Usually delivery. Voices drop in volume and clarity at the end of a long sentence. Two shorter dictations beat one long one, and our notes on speaking clearly for dictation cover the rest.
Honest limits
Voice Keyboard Pro types text. It does not drive Acrobat. There are no spoken commands for creating a highlight, exporting comments, applying a stamp or navigating pages, and that is deliberate rather than an omission. Voice control of an application interface is a different product with a different failure mode, and mixing the two means every stray word risks triggering an action in a document you are reviewing.
It needs a network connection, because transcription runs on fast cloud infrastructure rather than on your machine. It runs on macOS and iOS, so a Windows-only Acrobat workstation is out of scope. And dictation is audible, which is a real constraint in an open-plan office or a shared review room, and a real constraint when the comment you are speaking concerns a confidential matter.
On privacy: the server stores only operational pings. No audio and no transcript content is retained. That is a fact about our system, not legal advice about yours. If you review documents under an NDA, a protective order or a client confidentiality obligation, read our privacy policy yourself and clear it the same way you would clear any other tool that touches those documents. If you work in a regulated setting, our post on Mac dictation for lawyers covers that thinking in more depth.
Frequently asked questions
Does Adobe Acrobat have built-in dictation?
No. Acrobat has no voice input feature. You use a system-wide dictation tool, which types into Acrobat's comment boxes and form fields the same way a keyboard does.
Do I need an Acrobat plugin or extension?
No. Nothing is installed into Acrobat and nothing changes in its configuration. This is the main reason the approach works on managed and locked-down machines.
Does this work in Acrobat Reader as well as the paid versions?
Yes. Any text field you can type into is a field you can dictate into, in Reader or in the full product. What differs between them is which annotation tools you have, not how text gets into the box.
Can I dictate into a PDF form field?
Yes, and it works well for narrative fields. Type the identifiers rather than speaking them, because a wrong digit in an account number or a date reads as correct and survives every later check.
Will it work in other PDF apps?
Yes. Nothing here is Acrobat-specific. Preview, PDF Expert, Skim and the PDF readers built into reference managers all use standard text fields, so the same hotkey works in all of them.
Can I dictate in a language other than English?
Yes, and you can dictate in one language and have the text arrive in another, across 24 languages. For anything binding, have a fluent human read it before it goes out.
Is it faster in practice, or does the editing cancel out the gain?
For prose it is faster, and the reason is the size of the gap. Speech runs at roughly three times comfortable typing speed. Even with a correction pass, a spoken paragraph beats a typed one. The larger win is not speed anyway. It is that comments you would have compressed into one word get written out in full.
Try it on the next document that comes back for review
Pick a PDF you owe someone comments on. Read a section with the highlighter only, then go back and dictate a real comment on each highlight, speaking the reasoning and typing the references. Then read what you produced next to the comments you left on the last document.
The difference is not that you worked faster. It is that this time the reader knows what you meant. Voice Keyboard Pro has a free tier with daily limits, which is enough to test this on one real review. Pro is $4.99 a month or $34.99 a year if you decide the terse comments were never what you actually wanted to write.