AI in practice
Wispr Flow Review: My Daily Dictation Workflow
Wispr Flow works surprisingly well in German. The app instantly transforms spoken language into formatted text, automatically adding punctuation, paragraphs, and making numerous corrections.
I primarily use Wispr Flow for longer texts, emails, and instructions for AI tools. During my eight-week trial, the app processed 63,941 dictated words, averaging 177 words per minute. This is my personal result from the video, not a guaranteed product specification.
The app featured in my video is called Wispr Flow. Many people also search for “Whisper Flow,” “Wisper Flow,” “Wispr Flow,” or “Wisprflow.” The combined term “Whisperflow” may now refer to a different product. This article focuses exclusively on Wispr Flow from Wispr AI.
Does Wispr Flow Work in German? My Conclusion
For me, Wispr Flow is worthwhile because I speak my thoughts faster than I type. The advantage isn’t solely about speed; it’s that my spoken draft is directly transformed into usable text.
I tested all the common dictation tools, but I ultimately chose Wispr Flow.
The German language recognition works well in my daily use. I occasionally need to supplement the personal dictionary with names and technical terms. You should also carefully review the privacy settings before your first extended use. If you plan to dictate customer data, read the privacy section for businesses first.
You can start with the free version and test the difference yourself.
My Wispr Flow Tutorial as a Video
The video above shows the setup process directly on my Mac. I also demonstrate how I use Wispr Flow to dictate a blog post and build a small application in Cursor. The full transcript, with timestamps, is available at the end of this article.
The video was recorded on July 26, 2025. Some details have changed since then. There is now an Android app, new pricing plans, and updated names for the privacy settings. The following text reflects the status as of October 13, 2026.
What is Wispr Flow?
Wispr Flow is a dictation app from Wispr AI. You speak into your microphone, and the app inserts the text at the current cursor position. This works on your desktop in various text fields, such as email programs, documents, browsers, and AI tools.
The app doesn’t simply transcribe every spoken word. It removes many filler words, adds punctuation, and recognizes simple corrections. If I correct myself during dictation, Wispr Flow often automatically adopts the new phrasing.
According to the official language overview, Flow supports over 100 languages, including German and the “German - Swiss” variant, which uses “ss” instead of “ß.” Flow recognizes the language for each dictation, not each word. I recommend only activating the languages you actually use; this helps with automatic language detection.
In the video, I show a toggle for “Smart Formatting.” According to the current documentation, Smart Formatting is now enabled by default. On Mac and Windows, you can customize it under “Personalization” and “Style.” On iPhone, you can toggle it on or off in the settings. On Android, it runs automatically.
The app is developed by Wispr AI, Inc. The company was founded by Tanay Kothari and Sahaj Garg, according to the official media kit.
Setting Up Wispr Flow in German
The initial setup takes only a few minutes. On a Mac, Wispr Flow requires access to your microphone and accessibility features. Accessibility allows the app to insert the finished text into other programs. On Windows, the app only requests access to your microphone.
Here’s how I recommend getting started:
- Download Wispr Flow for your device.
- Allow microphone access.
- On a Mac, enable accessibility features for text input.
- Select German from the list of languages.
- Speak a short test text into an empty document.
- Add names and technical terms to the personal dictionary.
- Review the “Settings” and “Data and Privacy” to see what Wispr can store and use.
By default, on a Mac, you need to hold down the Fn key to speak. Pressing Fn + Spacebar activates hands-free mode. On Windows, use Ctrl + Windows Key and the spacebar for hands-free mode. You can customize these shortcuts, according to the shortcut documentation.
My Tip for Better German Texts
Speak in complete thoughts, but not artificially slowly. Short pauses between paragraphs are helpful. Adding names and technical terms to the dictionary is a good idea right away. For example, I have my name stored so that Flow writes “Artur” without the “h.” I have both German and English enabled as languages.
For example, I explicitly say “new paragraph” when I need a clear separation. Wispr Flow usually handles the rest of the formatting automatically. However, it’s still a good idea to review important texts before sending or publishing them.
Wispr Flow Download for Mac, Windows, iPhone, and Android
Wispr Flow is currently available on four platforms, each with slightly different requirements.
| Platform | Minimum Requirements | Important Note |
|---|---|---|
| Mac | macOS 12 or later | Supports both Apple Silicon and Intel processors. |
| Windows | Windows 10 or 11 | Only x64; not compatible with Windows on ARM. |
| iPhone | iOS 18.3 or later | Only compatible with iPhone; not iPad. |
| Android | Android 13 or later | Smartphone with an accelerometer. |
According to the Wispr Flow system requirements, iPads, Linux systems, Chromebooks, virtual machines, and remote desktop environments are not supported. Android tablets may work, but are not officially optimized.
On iPhone, you add Flow as a dedicated dictation keyboard. On Android, your regular keyboard remains active. A Flow Bubble appears above the keyboard when you open a text field. Password fields, banking apps, or specific input masks may restrict direct use. On all devices, an internet connection is required for dictation.
Is Wispr Flow Free? Pricing and Free Limits
Yes. The free plan is completely free of charge. In some help articles, it may still be referred to as “Basic.” The “Pro” plan costs US$15 per user per month, according to the official pricing page. With an annual payment, it’s US$12 per user per month.
| Plan | Price (as of October 13, 2026) | Included |
|---|---|---|
| Free (Mac or Windows) | Free | 2,000 words per week |
| Free (iPhone or Android) | Free | 1,000 words per week |
| Pro (Monthly) | US$15 per user per month | Unlimited dictation, Command Mode |
| Pro (Annual) | US$12 per user per month | Same as Pro, with annual billing |
| Growth | US$23 per user per month, US$18 with annual payment | Pro features plus SSO, admin controls, data processing agreement |
| Enterprise | Contact for pricing | Model training permanently disabled, user management, audit logs |
As of October 13, 2026, the pricing page no longer mentions a trial period. Previously, there was a 14-day Pro trial without requiring a credit card. Now, you can start directly with the free plan, and the pricing page indicates that you don’t need a credit card for that. Students and educators with a valid educational email address are eligible for a 50% discount on the Pro plan. However, it’s always a good idea to verify the details in the checkout process, as pricing may change.
For occasional use, 2,000 words per week may be sufficient. However, for my longer texts, that limit would be reached quickly. In the video, I mention that I dictate approximately 30,000 to 40,000 words per month. If you dictate regularly, the Pro plan is likely the more realistic option.
The Pro plan also includes Command Mode, which allows you to edit selected text using voice commands. This feature is disabled by default and can be enabled under Settings and Experimental. According to the Wispr help documentation, a paid Pro or Enterprise account is required.
My Experience with Wispr Flow
I’m not just using Wispr Flow for short messages. In my daily workflow, it’s particularly helpful for tasks that would otherwise require a significant amount of typing to create an initial draft.
Longer Texts and Blog Posts
In the video, I give Claude a general outline for a blog post. Then, I dictate my thoughts, examples, and the desired structure. The result isn’t automatically a finished article, but it’s a significantly better starting point than a blank document.
This workflow aligns well with my approach to self-reinforcing business systems. I first speak my thoughts, then refine them, and later use them as the basis for further content.
Emails and Messages
Short emails are particularly quick to compose. Wispr Flow often recognizes whether I’m writing in a casual or formal tone. However, I always double-check the recipient, numbers, and names before sending.
Programming with Cursor and Claude Code
In the video, I build a small Sudoku game in Cursor using voice commands. Instead of dictating individual lines of code, I describe the desired functionality and let the AI agent create the technical design.
That is where voice input is particularly valuable to me. I can explain requirements, changes, and errors more quickly. Wispr Flow thus becomes an input layer for other AI tools. For example, I dictate instructions to Claude Code in the terminal today.
In my articles about AI in Business, I demonstrate how I use AI tools in a business context. My complete toolset, including pricing, is detailed in AI Tools for Freelancers.
What Doesn’t Always Work Perfectly
Technical terms and unusual names sometimes require correction. In rare cases, a punctuation mark might not be correct. Similarly, sentences with rapidly changing languages can make recognition more difficult.
Wispr Flow also won’t do all your thinking for you. A poorly thought-out idea, dictated quickly, remains a poor idea. The app simply reduces the friction between an idea and text.
177 Words Per Minute: What That Number Really Means
After eight weeks, my statistics showed 63,941 words and an average of 177 words per minute. This number measures my usage of the app. It doesn’t mean that every user will always achieve that exact speed.
When typing normally, I have to input each letter myself. When dictating, I express a complete thought and let Flow handle the initial formatting. This saves me a significant amount of time, especially when working on longer drafts.
Therefore, for me, that number is a good personal indicator of how much dictation accelerates my writing process. How quickly you work depends on your speaking pace and the necessary post-processing.
Data Privacy for Business Owners: Cloud, USA, and Data Processing
The transcription process doesn’t run locally on your device. According to the official Data Controls, Wispr always processes audio in the cloud. Therefore, you can’t dictate without an internet connection. You should be aware of this before dictating confidential content.
Where Your Data Goes
According to the Privacy Policy dated August 19, 2026, Wispr transfers data to the USA and processes it there. Wispr also uses AI models from other providers for dictation. According to Wispr, it has agreements with all these providers that ensure no data is stored. These providers are not permitted to use your data for training purposes.
According to Wispr, the transfer is encrypted using TLS 1.2 or later, and stored data is encrypted using AES-256. Wispr also mentions a SOC-2 Type II report and an ISO-27001 certification, both provided by the auditing firm A-LIGN. You can find more information on the Privacy and Security page.
The Four Settings You Should Check
You can find these in the app under Settings and Data and Privacy:
- Improve the model for everyone: This setting used to be called Privacy Mode. If you disable it, Wispr will not use your audio, transcripts, and corrections to train its models.
- Dictation cloud storage: This determines whether Wispr stores your audio, transcripts, and history on its servers. If disabled, your dictation history remains on your device, and Wispr deletes the audio immediately after transcription.
- Context awareness: Flow will then read a limited amount of text from the active app to ensure that names and technical terms are correct. If disabled, Flow will not read anything from your screen.
- Local data storage: Here, you can specify whether your history should remain on your computer, be deleted after 24 hours, or not be stored at all.
Regardless of these settings, according to Data Controls, Flow always uses the name of the app and the surrounding text in the input field. This allows the app to choose the tone, capitalization, and punctuation. On the Growth and Enterprise team plans, model training is disabled by default and the setting is locked.
Data Processing and Customer Data
If you dictate personal data from customers, Wispr processes that data on your behalf. In such cases, you typically need a data processing agreement (DPA) as required by Article 28 of the GDPR. Wispr lists this agreement (MSA and DPA) in its pricing table starting with the Growth plan, not with the Free or Pro plans. In addition, data is transferred to the USA.
You should discuss whether this is suitable for your company with your data protection officer. I am not a lawyer and do not provide legal advice.
AI service providers themselves recommend: no passwords, no personal data, no customer data. I also share this rule in my courses. This rule also applies to Wispr Flow, since every dictation goes to the cloud. For prompts, your own drafts, and instructions to AI tools, you don’t need such data. For health data, access data, and internal documents, you should first check your own data protection obligations.
Therefore, I do not claim that Wispr Flow is suitable for any confidential content. If you need dictation that never leaves your device, consider local tools. You can find information about these in the section on alternatives.
Who Will Benefit from Wispr Flow?
Wispr Flow is a good fit for anyone who needs to convert a lot of thoughts into text. This is particularly true for freelancers, authors, consultants, developers, and people who regularly work with AI tools.
I would recommend testing the app if you:
- write many emails or long documents,
- prefer to explain prompts and tasks instead of typing them,
- often record thoughts on the go,
- want to maintain a custom dictionary of names and technical terms,
- are willing to check important texts afterward.
Wispr Flow may not be the best fit if you primarily write short, individual sentences. Also, if you handle highly confidential content, a cloud-based processing solution might conflict with your requirements. Users of Linux, iPad, Chromebook, or Windows on ARM are currently not officially supported.
Alternatives to Wispr Flow
I tested all the common dictation tools and ultimately chose Wispr Flow. However, Wispr Flow isn’t suitable for everyone. There are three scenarios where another tool might be a better choice:
- You want your voice data to never leave your device. In that case, Superwhisper or VoiceInk, with their local models, are worth considering.
- You prefer a one-time purchase rather than a monthly subscription. VoiceInk and MacWhisper are available as one-time purchases.
- You don’t want to install any software. macOS and Windows have a built-in dictation feature.
I compare all the tools in terms of pricing, platforms, and privacy on Wispr Flow Alternatives. You can find a direct comparison with the most well-known competitor on Wispr Flow vs. Superwhisper.
For conversations and brainstorming on the go, I also use the Plaud Note Pro. You can read about my experience with it on Plaud Note Pro Review.
My Recommendation
Start with a real-world test. Don’t just dictate a prepared example sentence. Write an email, a longer paragraph, and a prompt for your most frequently used AI tool.
After just a few minutes, you’ll notice whether the German speech recognition works well with your voice and your workflow. For me, Wispr Flow has become an integral part of my writing process.
You can find more articles on AI and business in this overview: AI in Business.