Free tool
Voice prompt builder
Speak into {{transcription}}, watch your prompt assemble live, copy into ChatGPT or Claude. Local dictation in your browser. No account.
Saved in your browser. Nothing uploaded to our servers
Saved in your browser. Nothing uploaded to our servers.
Speak the variable, copy the full prompt
Pick a template, hold the mic button, and dictate into the voice slot. The assembled prompt updates live, same hold to dictate flow as UBIK Flow.
Built-in voice demos plus templates from your AI Prompt Library.
Downloading decoder_model_merged_q4.onnx
Voice slot
Hold the button below and speak. Release to stop. Edit the text if needed.
{{transcription}}
Click to dictate
Listening… click the mic again to stop and transcribe.
Other fields
Type or paste supporting context. The voice slot is filled by dictation.
Extra context
Dictation runs locally in your browser. Nothing is uploaded to our servers.
Assembled prompt
Copy and paste as your first message in ChatGPT, Claude, or any AI chat.
Turn spoken notes into a polished email or Slack message.
About this tool
Speak the variable, copy the full prompt
Some parts of a prompt are easier to say than type: meeting recaps, standups, rough ideas, spoken notes for an email. The Voice prompt builder lets you pick a template, click the mic, dictate into the voice slot (usually {{transcription}}), and watch the assembled prompt update live before you copy it into ChatGPT or Claude.
Dictation runs locally in your browser. Your voice never uploads to UBIK Flow servers.
What this tool does
- Voice-first layout with a prominent mic slot and live assembled preview
- Three built-in voice demos always available (message, meeting recap, Slack standup)
- Templates from your library plus deep links from starter packs
- Dictate secondary fields too (same mic pattern as the AI prompt library runner)
- Local dictation with spoken language selector (EN / FR)
How it connects to the prompt cluster
| Tool | Role |
|---|---|
| AI prompt starter packs | Import role templates |
| AI prompt library | Save, edit, mark voice slot |
| Voice prompt builder | Speak {{transcription}}, preview full prompt |
| AI fluff stripper | Clean AI output before paste at work |
This is the browser preview of how UBIK Flow fills template variables by voice, without leaving your workflow.
Who this is for
- Managers and ICs who think out loud and hate typing long context
- Sales and CS dictating call notes into structured prompts
- Marketers turning spoken campaign ideas into brief-ready prompts
- Writers and operators who already use ChatGPT daily and want faster first messages
- Privacy-conscious teams who refuse cloud dictation for client or roadmap content
Tips for better voice prompts
- Pick the spoken language before you start (improves accuracy)
- Fill
{{context}}first when the template needs audience or channel constraints - Review the live preview before copy; fix names and numbers manually
- Run the AI fluff stripper on the AI reply before Slack or email
Privacy
- Speech-to-text runs entirely in your browser
- Templates and filled values stay in localStorage with the prompt library
- No account, no server-side transcript storage
Compare that to cloud dictation APIs that require sending audio over the network.
Related free tools
- AI prompt library: full template CRUD and per-field dictation
- Speech to text: quick local dictation without a template
- Live speech to text: pause-based streaming dictation
- Microphone test: verify input before an important dictation session
When to use UBIK Flow instead
This page shows voice-filled variables inside a template. UBIK Flow on Mac or Windows adds hold-to-dictate in any app with global shortcuts, larger on-device models, and AI editing where you already work (Slack, Gmail, Notion, IDEs).
Use this browser tool to try the pattern. Install UBIK Flow when voice prompts are daily habit and tab switching is the friction.
Questions
UBIK Flow
Like this, but everywhere you work
UBIK Flow is a desktop app that brings local speech recognition and AI into your daily workflow, on top of what you're already doing, without copy-paste.
- Dictate in any app, with on-device transcription
- AI sees your screen and understands your context
- ChatGPT and Claude in a side panel, one shortcut away
Mac and Windows. Works offline. Free download.
Free tools
More free tools
Microphone test
Check your mic with a live level meter, frequency view, device details and a practical quality rating. No audio is uploaded.
Speech to text
Record a thought, stop, and turn your speech into editable text. Free, with no account or audio upload.
Audio to text
Upload MP3, WAV, or M4A and transcribe locally in your browser with Whisper. Copy or download the text. Your audio never leaves this device.
Filler word counter
Count filler words in a transcript, see matches per 100 words, and review highlights in context. Paste text or import a document. Nothing is uploaded.