Dictation for journalists: file copy 5× faster
SpeechFlow lets reporters draft stories, field notes and social posts by voice on Mac — clean copy at the cursor, zero data retention. Free to start.
Journalism runs on deadlines — but your hands can only type so fast. Reporters average 40–60 words per minute at the keyboard; speaking flows at 150–180. SpeechFlow closes that gap: hold a key, speak your lede, release — and clean, punctuated copy appears right where your cursor sits, in any Mac app, with nothing stored anywhere.
The journalist's pain point
You get back from a press conference with forty minutes before the slot closes. You have quotes in your notebook, an angle in your head and stiff fingers from a cold morning outside. The bottleneck isn't thinking — it's transcribing your thoughts into words fast enough. Add daily email threads with sources, social posts, pitches to editors and reactive copy, and a working journalist is typing thousands of words a day. That wears on your wrists and eats time you don't have.
Existing solutions fall short. Apple's built-in dictation drops punctuation, leaves in every “um” and “so” and never adjusts tone. Cloud dictation tools upload your audio to someone else's server — a real concern when a source's identity or an unreleased story is in play.
How SpeechFlow fits the newsroom workflow
SpeechFlow is a native macOS app — Apple Silicon, about 50 MB, no Electron, no browser tab. The mechanic is simple:
- Click into your story draft in whatever app you use — your CMS, Word, Google Docs, Notes, even a Slack DM to your editor.
- Hold Control and speak naturally. Fillers, false starts, tone — don't worry about any of it yet.
- Release. A cleanup LLM strips the fillers, adds punctuation and adapts the tone, then inserts the finished text directly at your cursor.
There is no dictation window to manage and no copy-paste step. You keep your eyes on the page and your mind on the story. The full AI dictation guide explains the cleanup layer in more detail if you want to understand what the LLM actually does.
What journalists actually dictate
| Task | How dictation helps |
|---|---|
| Story first drafts | Speak the lede, the nut graf and the body structure while the reporting is fresh. Editing a spoken draft is faster than writing from blank. |
| Field notes | Dictate observations, atmosphere and direct quotes into Notes or a text file immediately after an interview — before memory fades. |
| Headlines & social posts | Try five headline variants by voice in under a minute; dictate tweet or LinkedIn copy without context-switching to a new app. |
| Emails to sources & editors | Voice-compose follow-ups, requests and pitch emails. See also: dictating emails on Mac. |
| Interview prep notes | Brain-dump your question list and background context by voice before a call. |
An honest word about transcription
SpeechFlow is a drafting and writing tool — it captures your own voice as you compose copy. It is not a court-grade interview transcription tool. If you need a verbatim transcript of a recorded interview with speaker labels and timestamps, a dedicated transcription service is the right choice. SpeechFlow's value is in helping you write faster, not in transcribing someone else's words for the record.
Privacy and source protection
Every journalist who covers sensitive topics has to think about what data leaves their machine and where it goes. SpeechFlow's position is simple: zero data retention. Your audio and text are processed and then gone — nothing is logged server-side.
If source protection is a serious concern, the BYOK (Bring Your Own Key) plan goes further. Your audio travels directly from your Mac to your own OpenAI, Gemini or Groq account — no SpeechFlow server is in the loop at all. BYOK is a one-time purchase of €69. For teams where reporters share a deployment, that's a straightforward security story to tell your editor-in-chief or IT department. More on the accuracy and privacy settings that affect what gets sent.
FAQ
Can I use SpeechFlow in my CMS or any other web-based tool?
Yes. SpeechFlow inserts text at the system cursor, so it works in any Mac app and any website open in a browser — your CMS, Google Docs, Basecamp, whatever your newsroom uses.
Is this a transcription tool for recorded interviews?
No — and it is worth being clear about that. SpeechFlow captures your own voice as you speak to compose copy. It is not designed to transcribe pre-recorded audio files or someone else's speech with speaker attribution. Use a dedicated transcription service for that.
How does SpeechFlow protect the identity of my sources?
SpeechFlow has zero data retention — nothing you dictate is logged or stored on a SpeechFlow server. On the BYOK plan your audio goes directly to the AI provider you control (OpenAI, Gemini or Groq), with no intermediate server. SpeechFlow makes no compliance certifications, but the data architecture is as minimal as it gets.
What does it cost?
Free for 2,500 words per week with no credit card required. Pro is €10/month or €70/year and removes the limit. BYOK is €69 once (lifetime) and routes audio through your own API key.
Does it work offline or on a Windows machine?
No — SpeechFlow is a native macOS app for Apple Silicon Macs and requires an internet connection for the cleanup LLM step.
Deadlines won't get easier — but filing copy can. Try SpeechFlow free and get your first 2,500 words this week with no card needed.