Wispr AI in talks for $260M Menlo Ventures-led round at $2B valuation as voice dictation moves toward 'voice OS'

Wispr AI, the company behind the voice-dictation app Wispr Flow, is in talks to raise about $260 million in a round led by Menlo Ventures at a valuation close to $2 billion, Bloomberg reported on May 12, 2026. The round has not closed and the final terms could move, but if it lands at these numbers it nearly triples Wispr’s $700M post-money from a $25M raise late last year.

Menlo Ventures led Wispr’s $30M Series A in June 2025; this would be the firm’s second lead position into the company in under 12 months. Notable Capital, which led the $25M follow-on last November, would presumably participate. Wispr Flow has crossed roughly 2.5 million downloads between late 2025 and early 2026, with enterprise adoption spreading across hundreds of organizations including several Fortune 500 names.

The company is positioning its next phase as a “voice OS” rather than a dictation app, branching from the speak-to-text core into broader voice-driven interaction across documents, chats, and productivity tools.

Why this matters

Voice input has been a tantalizing AI category for two years and has largely failed to graduate from feature to product. Apple’s Voice Control, Google’s voice typing, Dragon NaturallySpeaking, and the long tail of meeting-transcription tools (Otter, Fireflies, Granola, Read) have all built loyal user bases but have not produced a $2B-scale standalone company. Wispr’s argument is that the right product shape is not a transcription tool that lives inside Slack or Word; it is a system-level voice surface that captures speech anywhere on the OS and produces polished text or actions directly into the active app.

The market backdrop for this round:

• Whisper, GPT-4o-voice, Gemini Live, and Claude Voice have effectively commoditized the underlying transcription quality. Wispr is not winning on ASR accuracy; it is winning on workflow integration.
• xAI’s Grok TTS launches in March and Grok-4.3 voice in May put voice into a generalist consumer assistant. Wispr’s defensibility is the “productivity” rather than “assistant” framing, capture, format, and deposit text where the user is already working.
• Microsoft Copilot, ChatGPT, and Claude desktop apps all ship voice modes natively. Wispr’s pitch is that the OS-level keystroke-replacement layer beats any in-app voice button for daily-driver use.

The $2B valuation says VCs are willing to bet that the voice-input layer is durable as a standalone category, not a feature Microsoft or Apple absorb. That bet is non-obvious, Apple’s Voice Control and Microsoft’s Voice Access exist and are free, but the engagement data on Wispr Flow apparently supports it.

Buyer take

If you are a Wispr Flow user or considering deploying it across your team, three things to track:

• Enterprise SKU and admin controls. $2B-scale fundraises usually accelerate the enterprise motion. Expect SSO, audit logging, and tenant-isolation features to ship over the next two quarters. If your security team blocked Wispr earlier, re-evaluate after the new tier lands.
• Data residency for transcribed text. Voice capture flows speech through Wispr’s servers (or a vendor’s). Confirm where audio and transcripts are stored, how long they persist, and what training opt-outs exist.
• OS-level keystroke injection. Wispr’s product depends on injecting text into whatever app is focused. Some enterprise device-management policies block this. Pilot the deployment posture before signing.

For competing voice-input vendors (SuperWhisper, Talon Voice, Aqua Voice, voice-mode features in Notion / Linear / Granola), the Wispr round raises the bar on engineering and go-to-market spend. Expect feature parity and pricing moves over the next 90 days.

What is still unclear

The round has not closed publicly. Bloomberg’s reporting frames it as “in talks.” Wispr has not confirmed the lead, the round size, or the valuation. Pricing for any upcoming “voice OS” SKU is not disclosed. The breakdown between consumer-paid and enterprise revenue is not public, nor is the gross margin profile after speech-API costs.