Free Speech to Text Online — Dictate with Your Mic

By Muzamil Ahmad · Updated 2026-10-26 · Audio

#audio#speech to text#dictate online#voice to text#free speech to text#transcribe audio#mic dictation#free

# Free Speech to Text Online — Dictate with Your Mic

Typing is slow for a lot of people, and not just because of speed. Some people think out loud far better than they type. Some have a sore wrist, a busy pair of hands, or a deadline that does not leave room for careful keystrokes. Some are students who need to get lecture notes down quickly, or writers who lose an idea the moment they stop to find the right key. For all of them, the answer is often the same: say it instead of typing it.

ToolPilot Pro's free Speech to Text tool does exactly that. You open the page, choose a language, press a button, and speak into your microphone. Your words appear in a transcript box as you talk, and when you are done you can copy the text or download it as a plain text file. There is no account to create, no software to install, and nothing to pay. This guide walks you through the tool step by step, exactly as it appears on the page, explains how to get the clearest results from your microphone, and is honest about where the tool's limits lie.

If you have never dictated before, the first minute can feel a little strange. Most people get over that quickly, because the payoff is immediate: a paragraph that would have taken several minutes to type can be spoken in well under a minute, and the words on the screen usually look surprisingly close to what you meant. The trick is not to speak perfectly. It is to set the tool up correctly, speak at a natural pace, and then tidy the transcript afterwards — which the tool lets you do, because the transcript box is fully editable.

What the Tool Is, in Plain Terms

The Speech to Text tool is a single page with a small set of controls. It does not try to be a full transcription service, a recording studio, or a note-taking app. It does one job: it listens to your microphone and turns what you say into text that you can keep.

Here is what you will find on the page, from top to bottom:

That is the whole interface, and that simplicity is deliberate. There are no timelines to manage, no audio files to upload, and no settings buried in menus. You choose a language, you press a button, you speak, and you keep the text.

There is also one message you may see instead of the controls, and it is worth understanding before you start. If your browser does not support speech recognition, the page shows a clear notice: "Speech recognition is not available in this browser. It works in Chrome and Edge on desktop and Android. Please open this page in one of those browsers and try again — no transcript can be produced here otherwise." That is not an error you can work around by pressing buttons harder. It is the tool telling you plainly that it cannot work in that browser, rather than pretending to listen and giving you nothing. If you see that message, open the same page in Chrome or Edge and try again. This is covered in more detail in the honest-limits section later in this guide.

How the Tool Works

When you press "Start dictation", the tool asks your browser for permission to use your microphone, and then begins listening. As you speak, the browser's built-in speech recognition turns your speech into words and fills the transcript box with them, live, while you are still talking. You do not have to wait until the end of a sentence or a paragraph to see anything; the text grows as you speak, which makes it easy to spot a misheard word early and correct it later.

The language you chose in the dropdown matters here, because it tells the recognition what language to expect. If you are speaking English, pick English (US) or English (UK) according to your accent and spelling habits. If you are speaking Urdu, pick Urdu. If you switch languages mid-session, change the dropdown before you start a new stretch of dictation, because the tool will keep listening for the language you selected. Choosing the wrong language is one of the most common reasons a transcript looks wrong even when the microphone is fine: the tool is listening for the sounds of one language while you are speaking another.

When you press "Stop", listening ends and the transcript stays exactly where it is, in the box, ready for you to edit, copy, or download. Nothing is saved automatically to your device until you press "Download .txt", and nothing is saved to an account anywhere, because there is no account. If you close the tab without copying or downloading, the transcript is gone, so make that last step a habit: finish, tidy, copy or download, and only then close the page.

One point about how this tool differs from uploading a recording to a transcription website: there is no audio file involved at all. You are not recording a file, uploading it, and waiting for a service to process it. You are speaking live, and the words appear live. That keeps the workflow fast, and it is also the reason for the tool's privacy story, which is set out in full in the privacy section below: everything runs in the browser, and your audio and transcripts never leave your device.

Step-by-Step Guide: Dictate Your First Transcript

Step 1: Open the Tool in Chrome or Edge

Open the free Speech to Text tool in Chrome or Edge, on a desktop computer or an Android device. This is the single most important setup step, because speech recognition works in Chrome and Edge in practice, while other browsers may not support it at all. If you open the page in an unsupported browser, you will see the plain notice described above instead of working controls, and the only fix is to switch browsers.

Desktop Chrome or Edge is the most comfortable place to start, because you can see the full transcript box, edit it with a real keyboard, and download the text file easily. Android Chrome or Edge works too, and is a good option if you want to dictate on the move, but editing a long transcript on a phone screen is slower, so many people dictate on a phone and tidy the text later on a computer.

Step 2: Choose Your Language from the Dropdown

Before you press anything else, click the Language dropdown and choose the language you are about to speak. The eight options are English (US), English (UK), Spanish, French, German, Hindi, Urdu, and Arabic.

Take a second over the English choice, because it is the one people most often rush. English (US) and English (UK) are not just labels for the same thing; they tell the recognition which spelling and pronunciation patterns to expect. If you speak with a British accent and write "organise" rather than "organize", English (UK) is the better fit. If your accent and spelling are American, choose English (US). The difference will not fix a bad microphone, but it does reduce the small, annoying mishearings that add up over a long dictation.

If the language you want to speak is not in the list, the tool cannot listen for it. That is a real limit, and it is better to know it now than to dictate for ten minutes and wonder why the transcript looks like guesswork. Stick to the eight languages on offer, and pick the closest match for what you are actually going to say.

Step 3: Press "Start dictation" and Allow Microphone Access

Now press the "🎤 Start dictation" button. The first time you do this, your browser will ask for permission to use your microphone. You must allow it, or the tool cannot hear anything at all. Look for the permission prompt near the address bar, choose Allow, and speak a test sentence to confirm that words are appearing in the transcript box.

If no prompt appears and nothing happens, the most likely cause is that microphone access was blocked earlier — either for this site, or for the browser as a whole. In Chrome, you can check this in the browser's site settings for the microphone, and Google's own help page explains the steps for allowing and changing microphone access in Chrome. If you are on a work or school computer, your administrator may control microphone settings, in which case you may not be able to change them yourself. If you are using a headset, also check that its mute switch is off and that it is actually selected as the input device, because a muted or unselected headset is a very common silent failure: everything looks right, but no sound reaches the tool.

Once permission is granted, the "Stop" button becomes available, and the transcript box is live. You are now dictating. There is no countdown and no separate recording step — speaking is the whole action.

Step 4: Speak Naturally and Watch the Transcript Grow

Speak at your normal pace, in complete thoughts, and keep the microphone at a steady distance from your mouth. Do not shout, and do not whisper. A normal speaking voice, aimed roughly at the microphone, gives the cleanest result. Short pauses between sentences help the tool tell one sentence from the next, and they also help you: dictation works best when you think in sentences rather than in a continuous stream of words.

Watch the transcript box as you go, but do not try to correct every word while you are still speaking. That breaks your flow and usually makes the next sentence worse, because you have stopped thinking about what to say and started thinking about spelling. A better rhythm is to dictate a paragraph, press "Stop", read it back, fix the errors, and then start again for the next paragraph. The transcript box is editable, so corrections are easy, and editing after each paragraph keeps mistakes from piling up into a wall of text that feels too big to fix.

If a word keeps coming out wrong, try saying it more slowly and clearly, or try a simpler word that means the same thing. Proper names, technical terms, and unusual spellings are the words most likely to be misheard, and that is normal for speech recognition in general, not a fault in this tool specifically. Say the name once, check how it appears, and if it is wrong, type the correct spelling into the transcript box by hand. That manual fix takes two seconds and is far more reliable than repeating the same word five times and hoping the fifth attempt lands.

Step 5: Press "Stop" When You Finish a Passage

When you have finished the passage you want, press the "⏹ Stop" button. Listening ends, and the transcript remains in the box. Pressing "Stop" does not erase anything, and it does not download anything either; it simply ends the listening session so you can review and edit in peace.

It is worth pressing "Stop" deliberately rather than just trailing off and leaving the tool listening. A tool left listening will keep trying to turn background noise, a television in the next room, or a conversation nearby into words, and those stray words will land in your transcript. Ending the session cleanly keeps your text clean. If you want to dictate more, you can press "Start dictation" again and continue; the new words will be added to the transcript you already have, so you can build a long document in several passes.

Step 6: Edit the Transcript in the Box

This step is where a good transcript becomes a usable document. Click into the transcript box — remember, it is labelled "Transcript (live — you can edit it)" — and read through what you dictated. Fix misheard words, add punctuation where it is missing, split long run-on sentences, and delete any stray words that crept in from background noise.

Be honest with yourself about this step: speech recognition, even good speech recognition, does not produce a polished final draft on its own. It produces a very good first draft, and the difference between the two is editing. Budget a few minutes of editing for every ten minutes of dictation, and the result will read like something you wrote rather than something a machine guessed. Pay particular attention to numbers, dates, names, and words that sound alike — "their", "there", and "they're", for example — because those are the errors a quick skim tends to miss and a careful reader tends to notice.

Editing in the box has one more advantage: it keeps everything in one place. You are not copying a rough transcript into another app and then losing track of which version is current. The box is the current version until you copy or download it, so edit there, and only move the text on when you are happy with it.

Step 7: Copy the Transcript or Download It as a .txt File

When the transcript reads the way you want it to, you have two ways to keep it, and both are one click away.

Press "Copy transcript" to copy everything in the box to your clipboard. From there you can paste it into a document, an email, a notes app, or any other place you write. This is the best option when you plan to keep working on the text somewhere else, because pasting into your usual editor lets you format it, add headings, and save it in whatever format that app uses.

Press "Download .txt" to download the transcript as a plain text file. This is the best option when you want a simple, permanent copy that will open on any device, or when you want to keep the transcript as a record without pasting it anywhere yet. A plain text file has no formatting, no fonts, and no hidden extras, which is a strength for archiving: it will still open cleanly years from now, on any computer or phone.

Whichever you choose, do it before you close the tab. The tool does not save your transcript for you, and there is no account to recover it from later. Copy or download first, close second, and you will never lose a dictation to a closed tab.

Practical Tips: Getting the Clearest Transcript

Use the Best Microphone You Have

The single biggest factor in transcript quality is not the tool, and not your accent. It is the microphone. A headset microphone, or the microphone built into a decent laptop, will beat a phone held at arm's length or a laptop microphone buried under a pile of papers. If you have a headset, use it. It keeps the microphone at a fixed distance from your mouth, which matters more than most people expect: recognition struggles most when the volume of your voice keeps rising and falling as you move.

If you only have a built-in microphone, sit close to the device, face it, and keep the room as quiet as you can. Test with one sentence first — something ordinary, like a sentence about your day — and read it back. If it comes out clean, your setup is good. If it comes out garbled, change one thing at a time: move closer, turn down a fan, close a window, or switch to a headset. Changing everything at once makes it impossible to tell which change helped.

Find a Quiet Room, or at Least a Quieter Corner

Background noise is the second biggest factor, after microphone quality. A television, music, a washing machine, traffic through an open window, or another person talking nearby can all end up in your transcript as stray words, or can blur the words you actually said. You do not need a silent studio. A normal quiet room, with the door closed and the television off, is enough for most dictation. If you cannot get a quiet room, dictate in shorter passages and check each one before moving on, so a burst of noise only spoils a sentence or two rather than a whole page.

Be especially careful with fans, air conditioning, and heaters blowing directly across a microphone. That rushing sound is easy for a human ear to ignore and hard for recognition to ignore, and it can make an otherwise clear voice sound muffled. Moving the microphone a few centimetres out of the airflow, or turning the fan down for the duration of a dictation, often makes a visible difference.

Speak Clearly, but Speak Naturally

The best dictation voice is your normal voice, slightly slowed down, with clear word endings. Over-enunciating every syllable like a robot actually makes recognition worse, because it stops sounding like the natural speech the system was trained on. Equally, mumbling at full speed makes it worse in the other direction. Aim for the pace you would use when explaining something to a friend who is taking notes: steady, clear, and unhurried, with a brief pause at the end of each sentence.

Say punctuation only if it helps you, and remember that you can add it by hand afterwards. Some people like to say "full stop" or "comma" as they dictate, and that can work, but it is not required by this tool, and many people find it breaks their train of thought. The simpler habit is to pause where a sentence ends, dictate the next sentence, and then add punctuation during the editing step. Do whichever keeps you talking smoothly; the editing step exists precisely so that dictation does not have to be perfect.

Dictate in Paragraphs, Not in Marathons

Long, uninterrupted dictation sessions are where errors accumulate and attention drifts. A practical rhythm is to dictate one idea at a time — a paragraph, a bullet point, a single answer — then press "Stop", read it, fix it, and start the next one. This also protects you against accidents: if the browser tab crashes or the connection drops mid-marathon, an un-copied transcript can be lost, whereas a paragraph you have already copied or downloaded is safe.

Paragraph-by-paragraph dictation also improves the writing itself. Speaking a whole essay in one breath tends to produce long, winding sentences. Speaking one paragraph at a time, with a pause to read it back, tends to produce writing that is tighter and easier to follow. The pause is not wasted time; it is where the editing happens.

Example Uses: What People Actually Dictate With It

The tool is general-purpose, which means the best way to understand it is through the jobs people actually give it. The following examples are illustrative examples, not case studies or reports about real users, but each one maps to a real, everyday use of the Speech to Text tool.

Example: Drafting an Essay, Article, or Email by Voice

A student who struggles to start writing often finds that speaking the first draft is easier than typing it. As an example, imagine a student who needs a first draft of a short essay about a topic they already understand well. They open the Speech to Text tool, choose English (UK), and talk through their main points for ten minutes, one paragraph at a time, pressing "Stop" between paragraphs to tidy each one. At the end they have a rough draft in the transcript box, which they copy into their usual editor and rewrite properly. The dictation did not write the essay for them, but it got them past the blank page, which is often the hardest part.

The same pattern works for emails and messages. Dictate the message, read it back, fix the tone — a spoken draft often sounds too casual for a formal email, and the editing step is where you catch that — and then copy it into your email. Writers who use the site's other writing help, including the Best AI Writing Tools guide, often use dictation the same way: voice for the raw draft, careful editing for the finished piece. If most of your dictation is emails, the Best AI Email Assistants guide covers the wider email workflow, and a dictated first draft of a cover letter can be pasted into the free Resume Builder tool when you are ready to format it.

Example: Capturing Notes, Ideas, and Meeting Summaries

Ideas rarely arrive when you are sitting neatly at a keyboard. They arrive on a walk, in the car, or just after a meeting ends, when the details are still fresh and will be forgotten in an hour. Dictating a quick summary while it is fresh is one of the most practical uses of this tool. As an example, imagine someone who has just finished a study-group session and dictates a five-minute summary of what was agreed, who is doing what, and what still needs checking. They download the transcript as a .txt file, and the summary is saved before memory fades.

This pairs naturally with note-taking in general. If you want a broader look at how notes, voice, and writing tools fit together, the site's Best AI Note-Taking Tools guide is a useful companion read — dictation is the capture step, and a notes system is where the captured text lives afterwards. And if you need to check how long a dictated note has become, or whether it fits a word limit, paste the copied transcript into the free Word Counter tool, which counts words and characters for you. That two-step habit — dictate here, count there — is a small workflow that saves a surprising amount of guesswork. For the wider productivity picture — turning dictated notes into tasks and plans — see the Best AI Productivity Tools guide, and if you dictate your own summary after a meeting, the Best AI Meeting Assistants guide explains what dedicated meeting tools add on top of plain dictation.

Example: Writing When Typing Is Difficult or Painful

For some people, dictation is not a convenience; it is the only comfortable way to write. A sore wrist, a repetitive-strain problem, a broken arm, or a condition that makes typing slow can turn a short document into an exhausting task. Speaking the text instead removes that barrier, at least for the first draft. The editing step still needs a keyboard, but editing a transcript takes far fewer keystrokes than typing the whole thing from scratch, and that difference matters when every keystroke hurts.

The same applies to people who simply think faster than they type. If your fingers cannot keep up with your thoughts, ideas get simplified or lost in the gap between thinking and typing. Dictation closes that gap, because speaking is, for most people, much faster than typing. The transcript may need tidying, but the ideas arrive intact, which is the part that matters most.

Example: Dictating in Another Language

Because the Language dropdown includes Spanish, French, German, Hindi, Urdu, and Arabic as well as two forms of English, the tool is also useful for dictating in a language other than English. As an example, imagine a student practising Urdu who dictates a short paragraph, then reads the transcript back to check whether the words came out as intended. Or a French learner who dictates sentences they have just studied, as a way of testing pronunciation: if the tool hears the intended word, the pronunciation was probably close; if it hears a different word, that is a useful clue about which sound needs work.

This is a practice aid, not a language exam, and it should be used that way. The transcript tells you what the recognition heard, which is a good rough guide to clarity, but it is not a judgement of your accent or your ability. Use it to practise, not to grade yourself, and remember to change the dropdown to the language you are actually speaking before you start.

Example: Listening Back to What You Dictated

A useful trick, once you have a transcript you are happy with, is to listen to it read aloud. That is the reverse of this tool's job, and it is exactly what the site's Text to Speech tool does: it turns written text into spoken words. Reading a transcript silently and hearing it read aloud catch different errors. Your eyes skip over a repeated word or a missing word because your brain fills it in; your ears notice it immediately, because the spoken version stumbles. So a good finishing routine is to dictate with the Speech to Text tool, copy the transcript, and listen to it with the Text to Speech tool before you call it finished. The two tools are natural partners: one turns speech into text, the other turns text back into speech, and using them together gives you a simple read-back check that costs nothing.

Example: Dictating Scripts, Show Notes, and Study Notes

Creators use dictation the same way writers do, but for spoken content. As an illustrative example, imagine a YouTuber who talks through a video script outline while walking, then edits the transcript into a shooting script — the Best AI Tools for YouTubers guide covers the wider creator workflow around that habit. A podcaster can do the same for episode outlines and show notes; the Best AI Tools for Podcasters guide is the companion read there. And a student who dictates a summary straight after a lecture, while the material is still fresh, gets the same benefit as the study-group example above — the AI Tools for Students guide rounds out that study workflow.

Honest Limits: What This Tool Cannot Do

A free browser tool should be judged on what it honestly does, not on what a sales page wishes it did. Here are this tool's limits, stated plainly.

It works in Chrome and Edge in practice; other browsers may not support it. This is the biggest limit, and the tool does not hide it. If you open the page in a browser without speech-recognition support, you will see the notice quoted earlier in this guide, and no transcript can be produced there. That is not a bug you can fix with settings, and it is not a problem with your microphone. It is a browser-support limit, and the fix is to use Chrome or Edge on desktop or Android. The technical background for why this is a browser capability, rather than something any website can add on its own, is explained in MDN's documentation for the SpeechRecognition interface, which describes the browser feature that tools like this one rely on.

Accuracy depends on your microphone and on background noise. A good headset in a quiet room gives a clean transcript. A distant laptop microphone in a noisy room gives a messy one. Accents, fast speech, mumbling, overlapping voices, and unusual names and technical terms all reduce accuracy. The tool does not claim otherwise, and no free browser dictation tool can promise perfect accuracy in poor conditions. The practical answer is the one given in the tips section: improve the microphone and the room first, dictate in paragraphs, and edit afterwards.

It does not transcribe audio files, and it does not identify speakers. You cannot upload a recording and get a transcript back, and you cannot feed it a meeting recording and ask who said what. It listens live, to one microphone, and produces one transcript. If you need file transcription, speaker separation, or timestamps, you need a different kind of service, and you should expect to pay for it or to accept its own privacy trade-offs. This tool's simplicity is a strength for live dictation, and a limit for everything else.

It does not add punctuation, formatting, or structure for you. The transcript is plain text. Headings, bold text, bullet points, and paragraph breaks beyond the ones you create by pausing and editing are your job, in whatever editor you paste the text into. If you dictate a long, unbroken stream, you will get a long, unbroken transcript, and the editing step will be longer. Dictating in paragraphs, as recommended above, keeps that editing step manageable.

It does not save your work. There is no account, no cloud copy, and no history to return to. If you close the tab without copying or downloading, the transcript is gone. That is a deliberate privacy choice, and it is also a responsibility: copy or download before you close, every time.

It is not a substitute for careful writing. A dictated first draft is a first draft. Names, numbers, dates, and claims in it still need checking, just as they would in a typed first draft. If you are dictating something that will be published, submitted, or sent to an employer, read it back at least once, check the details, and do not assume that because it sounded right, it reads right.

Privacy: What Happens to Your Voice and Your Words

Privacy deserves its own section, because dictation involves your voice, and your voice can carry personal information — names, addresses, health details, work matters — that you would not want sent to a stranger's server.

For this tool, the position is simple and is stated on the tool page itself: recognition runs in your browser, and audio is not uploaded by this tool. Everything runs in the browser, and your audio and transcripts never leave your device. There is no account to link a transcript to, no server storage to breach, and no copy of your words sitting on a remote machine after you close the tab.

That does not mean you should dictate anything, anywhere, without thinking. A transcript you copy into an email, paste into a shared document, or download onto a shared computer is no longer private — it is wherever you put it. And a microphone in a shared room can be overheard by people nearby, just as a spoken phone call can. So the sensible habits are the same ones that apply to any writing: dictate personal or sensitive material in private, check where a copied or downloaded transcript is going, and do not leave a downloaded .txt file on a computer other people use.

One more honest note: the browser itself still needs microphone permission, and you should grant that permission only to sites you trust, for the time you need it. If you allowed microphone access and later want to remove it, Chrome lets you change a site's microphone permission in its site settings, and Google's help page on using your camera and microphone in Chrome explains how to allow, change, or block that access. Revoking permission after a dictation session is a good habit if you share the computer or simply prefer to keep permissions tight.

Troubleshooting: When the Transcript Is Wrong or Missing

Most dictation problems have one of a handful of causes, and they are worth working through in order before assuming the tool is broken.

Nothing appears when you speak. First, check that you are in Chrome or Edge and that you do not see the unsupported-browser notice. Then check microphone permission: if the browser never asked, or you previously chose Block, the tool cannot hear you, and you will need to allow the microphone for the site in the browser's settings. Then check the hardware: is a headset muted, is the correct microphone selected, is another app holding the microphone exclusively? Close other calling or recording apps and try again with a single test sentence.

The transcript appears, but it is mostly wrong. Check the Language dropdown first. Dictating Urdu while English (US) is selected, or French while German is selected, will produce confident-looking nonsense. Change the dropdown to the language you are speaking and start again. If the language is right, check the room and the microphone next: move closer, reduce background noise, and try a headset if you have one. If only certain words are wrong — names, technical terms, numbers — that is normal, and the fix is to type those words in by hand during editing rather than repeating them.

Words appear, then stop appearing. Press "Stop" and then "Start dictation" again. Listening sessions can end on their own, especially after a pause, and restarting is the normal way to continue. Dictating in paragraphs, with a "Stop" and a restart between them, avoids this problem almost entirely, because you are never relying on one very long session.

The transcript looks right, but the download or copy did not work. Check the transcript box first: is the text actually there? If it is, try the other option — if "Copy transcript" seems to have failed, use "Download .txt" instead, and vice versa. If you are on a phone, remember that clipboard and download behaviour varies by device, and that a desktop browser is the more reliable place for the final copy or download step.

The page shows the unsupported-browser message. Switch to Chrome or Edge on desktop or Android and open the Speech to Text tool there. There is no setting in the unsupported browser that will make the tool work, and trying to force it will only waste time. This is the tool being honest rather than broken, and switching browsers is the whole fix.

This Tool vs Paid Transcription Services

It helps to place a free live-dictation tool alongside the paid alternatives, fairly.

Paid transcription services usually accept uploaded audio or video files, return transcripts with timestamps and sometimes speaker labels, and handle long recordings that would be impractical to dictate live. Some also offer human review for higher accuracy. What they cost is money, and often privacy: your audio file goes to their servers, is stored there, and may be processed by people or systems you cannot see. For a recorded interview, a lecture, or a meeting that already happened, that trade can be worth it.

The Speech to Text tool is the opposite trade. It costs nothing, keeps everything in your browser, and gives you text immediately, but it only works live, only in Chrome and Edge in practice, only in its eight languages, and only as plain text that you edit yourself. For drafting, note-taking, brainstorming aloud, and getting words down when typing is hard, that trade is excellent. For transcribing an hour-long recording you already made, it is the wrong tool, and saying so is better than pretending otherwise.

If you are weighing voice tools more broadly — including tools that generate voices rather than transcribe them — the site's Best AI Voice Generators guide covers that side of the audio world, and it is a useful companion to this guide: transcription turns your voice into text, voice generation turns text into a voice, and knowing which direction you need saves a lot of searching.

Frequently Asked Questions

Is the Speech to Text tool really free?

Yes. There is no account, no sign-up, no trial, and no charge for copying or downloading your transcript. Open the free Speech to Text tool, dictate, and keep the text. That is the whole offer.

Which browsers does it work in?

In practice, Chrome and Edge, on desktop and Android. Other browsers may not support speech recognition, and if yours does not, the page will say so plainly instead of pretending to listen. If you see that message, open the page in Chrome or Edge and try again.

Which languages can I dictate in?

Eight: English (US), English (UK), Spanish, French, German, Hindi, Urdu, and Arabic. Choose the language from the Language dropdown before you press "Start dictation", and choose the one you are actually going to speak, not the one you wish you were speaking. The wrong language choice is the most common cause of a transcript that looks like guesswork.

Is my voice or my transcript uploaded anywhere?

No. Recognition runs in your browser, and audio is not uploaded by this tool. Your audio and transcripts never leave your device while you are using the tool. Once you copy or download a transcript, it goes wherever you paste or save it, so keep that destination in mind for sensitive material.

Can I edit the transcript before I save it?

Yes, and you should. The transcript box is labelled "Transcript (live — you can edit it)" for a reason: click into it, fix misheard words, add punctuation, and split long sentences before you press "Copy transcript" or "Download .txt". Editing is part of the workflow, not an admission that dictation failed.

What do I get when I download?

A plain text (.txt) file containing your transcript. It has no formatting, which means it opens on any device and stays readable indefinitely. If you need formatting, headings, or a styled document, paste the copied transcript into your usual editor and format it there — and if you want to check its length first, the free Word Counter tool will count its words and characters for you.

Does it work on mobile?

Yes, in Chrome or Edge on Android. Dictating on a phone works well for short notes and ideas, but editing a long transcript and managing the download are more comfortable on a desktop or laptop. A practical split is to dictate on the move and tidy the text later on a computer.

What should I do if the transcript keeps mishearing the same word?

Type that word in by hand during editing. Proper names, technical terms, and unusual spellings are the hardest words for any speech recognition, and a two-second manual fix beats repeating the word five times. If many words are wrong, check the Language dropdown and your microphone setup first, because those two causes explain most widespread errors.

Try It: One Minute of Dictation

You do not need to prepare, practise, or buy anything to find out whether dictation suits you. Open the free Speech to Text tool in Chrome or Edge, choose your language from the dropdown, press "🎤 Start dictation", allow microphone access, and speak one paragraph about anything — what you did today, an idea you have been turning over, the first paragraph of something you need to write. Press "⏹ Stop", read the transcript back, fix anything that needs fixing, and press "Copy transcript" or "Download .txt" to keep it.

If that paragraph came out close to what you meant, you have just found a faster way to write a first draft. If it did not, check the language, move closer to the microphone, find a quieter corner, and try once more. Dictation is a skill that improves with a little practice, and the first minute is the cheapest possible test. Your voice is already a writing tool — this page simply gives it somewhere to land, as text, ready to edit, copy, or download.

Reader reviews

–
Loading reviews…