An AI voice input app developed by Wispr in the United States. Just speak into the microphone, and the AI removes fillers such as “um,” adds punctuation and line breaks, turns your speech into natural prose, and writes it directly into whatever app you currently have open. Whether it is email, chat, or a code editor, you can use it without going through a dedicated transcription window. The official site claims roughly 4x the speed of typing (around 220 words per minute) and support for more than 100 languages. It runs on Mac, Windows, iPhone, and Android, and as of 2026 it also includes Notetaker for meeting transcription and an MCP-based integration with Claude and ChatGPT.
Key Features
- Voice input directly in any app: Rather than transcribing into a dedicated editor and asking you to copy and paste, it writes straight into the input field where your cursor is. Email, Slack, documents, terminals, code editors ─ almost anywhere a keyboard works, the same action works
- Automatic conversion from spoken to written style: The AI removes fillers, adds punctuation, and applies your mid-sentence corrections. If you say “actually, change that part to this” while speaking, the result reflects it
- Vocabulary and style learning: It automatically picks up proper nouns, personal names, and industry jargon, and you can also register terms in a dictionary by hand. Beyond that, it can adapt tone per app, so chat and email do not come out in the same register
- 100+ languages with automatic detection: It supports more than 100 languages including Japanese and detects which one you started speaking in. You can also lock the language manually
- Snippets (boilerplate by voice): Register frequently used greetings, signatures, or standard replies as snippets and expand them with a short phrase
- Notetaker and MCP integration: Notetaker records meeting transcripts with speaker identification, and it connects to your calendar and Slack. Referencing meeting notes from Claude or ChatGPT over MCP is a supported use case
- Security certifications: The official site states SOC 2 Type II, ISO 27001, and HIPAA readiness, along with a commitment that data is never sold and that data-sharing preferences stay under user control
Pricing
| Plan | Monthly | Billed annually | What’s included |
|---|---|---|---|
| Free | $0 | ─ | Up to 2,000 words per week on desktop, 1,000 words per week on iPhone, unlimited on Android. Voice input in 100+ languages, Notetaker (with a weekly meeting limit) |
| Pro | $15 / user | Effectively $12 / user (20% off when billed yearly) | Unlimited dictation, extended meeting retention, access to advanced AI models, early access to new features, prioritized support, shared dictionary and snippets, team billing and usage dashboards |
| Enterprise | Contact sales | ─ | Everything in Pro, plus SOC 2 Type II / ISO 27001 / HIPAA-ready BAA, SSO and SAML, audit logs, domain management, dedicated support, and volume discounts |
Pricing is as of August 2026. Check the official site for the latest information.
Pros & Cons
✅ Pros
- You can dictate on the spot without switching apps, which removes the copy-and-paste round trip
- Casual speech still comes out as clean prose, so there is little cleanup afterward
- It handles many languages including Japanese, and you can move between them
- For long emails, meeting notes, and first drafts, it is clearly faster than typing in many situations
- A free plan exists, so you can test the fit within the 2,000-word weekly allowance before paying
⚠️ Cons
- The free plan’s 2,000 words per week runs out quickly if you write regularly
- Because it is voice input, you need a quiet environment or one where speaking aloud is acceptable
- Proper nouns and technical terms are prone to misrecognition out of the box, so some dictionary tuning is required
- Since the AI reshapes your text, the exact wording can shift from what you said, which matters for precise quotations or fixed-format documents
- The core features assume cloud processing; this is not a product designed for fully offline use
Comparison with Similar Services
| Criteria | Wispr Flow | superwhisper | MacWhisper | Built-in OS dictation (macOS / Windows) |
|---|---|---|---|---|
| Provider | Wispr (US) | Superwhisper | Good Snippets | Apple / Microsoft |
| Platforms | Mac / Windows / iOS / Android | Mac / iOS | Mac | Bundled with each OS |
| Types directly into any app | Yes | Yes | Yes | Yes |
| AI style cleanup | Built in (filler removal, formatting) | Available via modes | Available via modes | Almost none (raw transcript) |
| Local processing | Mainly cloud-based | Local mode available | Mainly local | Partly local depending on device |
| Meeting transcription | Notetaker included | ─ | Audio file transcription | ─ |
| Pricing model | Free plan + subscription | Free plan + paid | Mainly one-time purchase | Free (bundled with OS) |
Who Is It For
- People who write a substantial amount of email, chat, and documents every day and find typing speed to be the bottleneck
- Writers and planners who want to capture what they say while thinking, already in readable form
- People who want to reduce keyboard use because of strain on their hands or shoulders
- People who move between English and Japanese while writing
- Teams that want meeting records and the writing that follows them in a single tool
Summary
Wispr Flow does more than turn speech into text: it takes what you said and shapes it into prose you can submit as is. Because the step of transcribing into a separate app and pasting disappears, it is easy to adopt as your everyday input method. On the other hand, the free plan’s 2,000 words per week is short for serious use, and at $15 per month ($12 effectively when billed yearly) it is not cheap for a text input tool. Start with the free plan, check how well it fits your speaking style and how accurately it handles your proper nouns, and decide from there.