Think the sentence, say it, see it correctly spelled. StarWhisper turns voice into text in any Windows app, runs locally on the device, and gives you 500 words a day on the free plan.
Decouple thinking from spelling, so ideas land on the page intact
Speak the sentence, see the correctly spelled words appear. Whisper handles spelling and punctuation, so your attention stays on what you want to say.
"Their," "there," and "they're" usually come out right because Whisper uses sentence context, not single-word guessing.
Google Docs, Microsoft Word, Outlook, Notion, any learning management system. One hotkey, any text field, no per-app setup.
Local Mode keeps audio on the device. No cloud upload, no data-residency concern, easier conversation with school IT.
500 words/day and 3,500/week on the free plan, no credit card. Enough for daily homework and short essays.
Strong support for English, German, Spanish, French, and dozens more. Dictate in the language of the assignment.
For a dyslexic writer the act of writing carries an invisible tax that most people never see. Every word demands a small decision: spelling, letter order, capitalization, homophone choice. Each decision is fast for a non-dyslexic brain and slow for a dyslexic one, and the cost compounds across a paragraph. By the time the sentence is on the page, the thought that started it has often shifted, been forgotten, or been narrowed down to whichever words were easier to spell. The finished writing reflects the spelling constraint as much as it reflects the original idea.
Voice-to-text removes that constraint at the source. You think the sentence, you say it, the words appear on the page already spelled. StarWhisper uses OpenAI Whisper, which handles spelling and punctuation as part of the transcription, so the writer never has to context-switch into "how do I spell that?" mode. The cognitive bandwidth that was going to spelling now goes to thinking about what you actually want to say.
The practical difference is that dyslexic writers using dictation can produce longer, more complex writing in less time, with fewer surface errors, and crucially with less of the emotional cost that makes writing feel like an ordeal. The page stops being a battleground of spelling decisions and becomes what it should be: a place to record thoughts.
Older speech engines processed audio one word at a time and made guesses without much context. Whisper does the opposite: it processes audio in chunks and uses surrounding words to choose the right form. That changes the dyslexic experience in three concrete ways.
First, homophones almost always come out right. "Their book is over there because they're done with it" gets transcribed correctly because the sentence context makes each form obvious. For a dyslexic writer this removes one of the most exhausting parts of self-editing, the constant second-guessing of homophone choice. Second, capitalization and punctuation get inferred automatically. Periods at the end of sentences, commas around dependent clauses, capital letters at the start of sentences, names, and most proper nouns. You can still go back and fix anything that came out wrong, but the default is usable. Third, common spelling traps just disappear. Words you have learned to avoid because you cannot reliably spell them become available again, because you are no longer the one spelling them.
Whisper is not perfect. Specialized vocabulary, rare proper nouns, and field-specific jargon can come out phonetically and need a quick fix. The workflow is to dictate the whole passage, then do a short proofread pass for any specialized words. For most school and college writing the proofread pass takes a fraction of the time the typing pass would have taken.
Dictation is one of the most commonly approved assistive-technology accommodations in IEP and 504 plans in the United States, and similar frameworks exist in most other countries. The principle is the same everywhere: a student with dyslexia is held to the same content standard as their peers, and dictation removes a mechanical barrier (spelling and motor production) that is unrelated to the content the student is being graded on.
StarWhisper is a general-purpose Windows dictation tool rather than a school-AT-vendor product, but the underlying capability is the same category that AT coordinators routinely approve. Schools that already permit dictation as an accommodation can typically extend that allowance to a free general-purpose tool like StarWhisper. The Local Mode architecture is often the clearest sell to school IT: audio stays on the device, no third-party cloud upload, no data leaving the school network. The privacy and offline mode page goes into the architecture in more detail for IT review.
For high-stakes proctored exams (SAT, ACT, AP, GRE) the testing body usually requires its own approved assistive software in its own controlled environment, so StarWhisper is not the right tool there. For everyday classwork, homework, essays, journals, and reading responses, StarWhisper is a viable tool that the student can use on a personal Windows laptop or any school-issued Windows device the school IT allows installations on.
The biggest single reason cloud-based dictation tools get blocked from school networks is the upload of audio data, especially for minors. School IT teams have to answer hard questions about what leaves the network, where it is stored, who can subpoena it, and how long it sticks around. A cloud dictation tool that uploads every audio clip to a third-party server is a difficult conversation, and "we send your child's voice recordings to a remote server" rarely gets enthusiastic approval.
StarWhisper sidesteps the entire conversation in Local Mode. The Whisper model is on the laptop. The audio is on the laptop. The transcript is on the laptop. Nothing leaves the network. From the school IT perspective the app is closer to a local spell-checker than to a cloud service. There is no account, no upload, no SaaS contract to review. Cloud Mode exists as an opt-in for users who want OpenAI's hosted accuracy, but it is not the default and can be disabled entirely.
For home use the same architecture means parents stay in control of what is on the device and what leaves it. Voice clips of a child's homework do not get added to a cloud vendor's training set; they exist for the duration of the transcription and then they are gone.
Open the document, focus the cursor, press the dictation hotkey, talk through the answer or the paragraph the same way you would explain it out loud. Stop dictation. Read what you wrote. Fix specific words that came out wrong. Restart dictation to add more. The friction is dramatically lower than typing because the spelling decisions are absorbed by the tool. The free tier's 500 words per day covers most middle-school and high-school homework volumes.
Dyslexic students often have rich verbal responses to reading and a much harder time writing them down. Dictation closes the gap. Talk through the response the way you would in a classroom discussion, watch it land as text. The result is usually richer than the same student would produce by typing, because the limiting factor has shifted from spelling to thinking.
Communication that involves writing is often where dyslexia is most visible to the outside world. A typed email betrays spelling patterns; a dictated email reads like the student's actual voice. For older students and adults, this directly affects how they are perceived in school and at work.
Dictate a summary of what you just read, what was said in class, or what you understood about a topic. The summary becomes a study guide the student can actually use. The act of saying it out loud also reinforces the concept in memory, which is a separate benefit of voice-based study workflows. For more on student-specific patterns see the voice to text for students page.
Dictation only works as an accommodation if the student actually uses it. Tools that require a paid subscription before any dictation works often do not survive contact with daily homework volume. StarWhisper's free plan is genuinely free: 500 words per day, 3,500 per week, no credit card, no expiring trial, no nag screens.
For most dyslexic students the free tier covers a normal school day. Pro is $10 per month or $80 per year for unlimited dictation, useful for heavy writing periods like research papers, college applications, or for adults using dictation as a primary writing tool at work. A 7-day full-access trial lets you find out whether the free cap is enough before committing.
The recommendation for most families is: install on the student's Windows device, use the free plan for two weeks, see whether the daily cap is comfortably above the actual writing volume, and only upgrade if dictation has clearly become a daily tool. Related pages: voice to text for ADHD for students whose accommodations overlap, and voice to text for writers for adults using dictation as a primary writing tool.
Other ways people lean on voice to take pressure off writing
Skip the working-memory wall and get ideas out at the pace of thought.
Essays, notes, study guides, dictated and edited in any Windows app.
First drafts, blog posts, novels, dictated at the natural pace of thought.
How Local Mode keeps audio on the device, for school IT and parents.