An AI dictation app that turns natural speech into clean, formatted text in any application — email, docs, chat, code editors — trimming filler and fixing formatting as you talk, across Mac, Windows, iOS, and Android.
Last verified · 2026-08-17 · by Moe Ameen
Wispr Flow (wisprflow.ai) is an AI dictation tool from Wispr, a San Francisco company founded in 2021 by Stanford classmates Tanay Kothari and Sahaj Garg. You press a shortcut, talk, and Flow inserts cleaned-up text into whatever field you were in — an email, a Google Doc, a Slack message, a code editor — editing out "um"s and false starts and applying basic formatting as it transcribes. The pitch is speed: most people talk far faster than they type, and Flow is built to make dictated text good enough to keep without heavy cleanup.
It runs on Mac, Windows, iOS, and Android (the Android app arrived in February 2026), with a Mac-first meeting note-taker that produces transcripts, summaries, and action items. The company says it is rolling out a new speech model, Canto, aimed at lowering transcription error rates, and it has expanded into meeting-notes territory alongside tools like Granola and Fireflies. In August 2026 Wispr raised a $280 million Series B at a $2 billion valuation, and says revenue has grown more than 150% for four straight quarters — context on why the category is heating up rather than a spec you need.
Pricing, verified on the site: a free tier with a weekly word allowance (reported around 2,000 words/week on desktop and 1,000/week on iPhone, with Android unlimited), Flow Pro at $15/user/month or $12/user/month billed annually for unlimited dictation and longer retention, and a custom Enterprise plan with admin and compliance controls. Confirm current allowances and prices on wisprflow.ai/pricing, since tiers change.
The honest boundary: Wispr Flow is an input tool. It is very good at getting spoken words onto the screen as usable text, and it stops there. It does not generate video, images, carousels, or persona footage; it does not caption or reframe media per platform; and it does not schedule or publish anything. Turning what you dictate into finished, distributed content is a separate job.
Wispr Flow solves the input half of content: it makes talking as fast as thinking. What it does not do is turn that talk into anything a feed can use. That is where [Kompozy](/) picks up. Treat Flow as the microphone and Kompozy as the studio behind it. Dictate the script for a talking-head video into Flow, paste it into Kompozy, and generate a [Persona Short](/glossary/persona-shorts) or [Persona Frames](/glossary/persona-frames) clip your face-locked avatar reads to camera — captioned and sized for 9:16. Dictate a rough outline of an idea and Kompozy expands it into a Blog Article, a brand-exact [Carousel](/glossary/hyperframes), [Quote Graphics](/glossary/output-buckets), and native Text Posts, all in one governed voice via the [Persona Brief](/glossary/persona-brief). Where Flow gives you fast words, Kompozy gives you the video, images, and persona formats a dictation app can't make — then publishes the whole set across the eight social platforms plus blog and email on [Autopilot](/glossary/autopilot).
The workflow compounds for anyone who thinks out loud. A five-minute spoken riff dictated in Flow — messy, tangential, human — is exactly the raw input Kompozy is good at atomizing: it pulls the throughline into a short, cuts the best lines into quote cards, and reshapes the whole thing into a week of posts. You never touch a keyboard to draft and never touch a design tool to finish. Flow removes the typing; Kompozy removes everything between a rough idea and a published, on-brand content calendar.
Wispr Flow is an AI dictation app that lets you speak into any application and inserts cleaned-up, formatted text — trimming filler and false starts as it transcribes. It runs on Mac, Windows, iOS, and Android, and includes a meeting note-taker that produces transcripts, summaries, and action items. It is made by Wispr, which raised a $280M Series B at a $2B valuation in August 2026.
Wispr Flow has a free tier with a weekly word allowance (reported around 2,000 words/week on desktop and 1,000/week on iPhone, with Android unlimited), a Flow Pro plan at $15/user/month or $12/user/month billed annually for unlimited dictation, and a custom Enterprise plan. Confirm current allowances and prices on wisprflow.ai/pricing, since the tiers change.
Wispr Flow runs on Mac, Windows, iOS, and Android; the Android app launched in February 2026. Its meeting note-taker has been Mac-first at the time of writing, so check wisprflow.ai for current cross-platform support before relying on it on a specific device.
No. Flow is an input tool — it produces clean text, not media, and it does not caption, reframe, schedule, or publish. To turn dictated text into carousels, quote cards, persona video, blogs, and scheduled posts, you pair it with a content engine. Kompozy is built for that step: paste Flow's output in and it generates and publishes finished content across nine platforms.
It is a strong input layer if you think faster than you type — scripts, hooks, captions, and outlines start as speech instead of a blank page. But it is only the capture half. Creators typically pair Flow with a generation-and-publishing engine like Kompozy so the fast dictated input becomes a finished, distributed multi-format week rather than one clean paragraph.