A round that reprices dictation as a platform bet
Wispr Flow closed a 280 million dollar Series B on August 17, 2026, at a 2 billion dollar valuation. Nine months earlier, in November 2025, the company was valued at 700 million dollars. Menlo Ventures led both rounds. New backers this time include Acrew, Forerunner, Goodwater, Peak XV, Together Fund and PLUS Capital; existing investors Notable Capital, NEA, Neo Ventures, 8VC and MVP Ventures all increased their commitments rather than being diluted out. Total funding raised now stands at 361 million dollars.
A near-tripling in nine months for a voice-dictation app is not typical even by AI-funding standards. It signals that investors are no longer pricing Wispr as a niche typing tool, but as a bet on who owns the voice layer of daily work software.
What the money is actually building
The company just launched Canto, a new speech recognition model it says cuts word error rates from around 30 percent to under 10 percent, the accuracy gap that has kept some professional users away from voice input entirely. Alongside Canto, Wispr shipped a meeting note-taking feature that records, summarizes and extracts action items, putting it in direct competition with dedicated note-taking tools like Granola and Fireflies rather than staying a single-purpose dictation app.
A new unit called Wispr Interface Labs, led by former Amazon Alexa engineer Ariya Rastrow, is exploring voice input beyond the phone screen, and the company already has a hardware partnership enabling quiet dictation on wearable devices such as the Oasis ring. Go-to-market teams have expanded into India and the United Kingdom since November.
Dictation tools do not stay single-purpose for long
The pattern is familiar from every category that started as a single feature and grew into infrastructure: a tool bought to solve one narrow problem, typing by voice, gradually becomes the default layer underneath several workflows at once, from drafting emails to summarizing meetings. Once a team's meeting notes, action items and dictated drafts all run through the same vendor, switching later means migrating three habits instead of one.
None of this is disqualifying. It is a reason to treat the accuracy claim as a number to verify internally rather than accept on the vendor's word, and to decide deliberately whether the meeting-notes feature is wanted before it becomes the default simply because it shipped inside a tool already installed.
The angle a European IT buyer should not skip
Wispr's expansion into the United Kingdom puts a voice-data processor squarely inside a market with its own GDPR-adjacent UK data protection regime; dictated speech, and now recorded meeting audio, counts as personal data the moment it captures a name, a client detail or a health reference spoken aloud in a call. Any procurement sign-off for the meeting-notes feature specifically, not just the original dictation tool, should confirm where that audio and its transcript are processed and stored, since a feature added after the original vendor review does not automatically inherit the same data-protection scrutiny.
For a company already running Wispr for dictation, the practical task this week is small: check whether the meeting note-taking feature is enabled by default, and route it through the same data-protection review the original tool received.
Read next: Brussels Fund Doubles Lovable to $13.3 Billion | Omilia Raises $67M as Banks Trust AI With Live Calls



