Documentation
How to use VoiceCraft's privacy-first tools. All processing runs in your browser where possible.
Privacy & architecture
VoiceCraft is built by System Logic with a "Local-First" architecture. For the main tools, your audio and text are processed directly in your browser using WebAssembly and modern web APIs. Data is not sent to our servers for transcription or text-to-speech unless you use a feature that explicitly requires it.
- Voice Recorder & transcription: Recording and AI transcription run in your browser. Audio stays on your device; we do not upload it for processing.
- Text-to-speech (Voice Generator): Synthesis runs in your browser where supported. No server round-trip for the main flow.
- Script Editor, Summarizer, Plagiarism Remover, and other text tools: AI runs locally in your browser. Your text is not sent to our servers for these operations.
- Data retention: We do not store your recordings, transcripts, or generated content on our servers unless you sign in and choose to save history. Session data in the browser may be cleared when you clear site data or use a private window.
- Security: All traffic to our site is encrypted (HTTPS). Account data is stored securely via Supabase.
Tool guides
Voice Recorder
Record audio directly in your browser with no time limit. Step-by-step:
- Open the Voice Recorder tool and allow microphone access when the browser prompts you.
- Click the record button to start; click again to stop. You can pause and resume during a session.
- Use "Live Transcription" to see real-time text as you speak, or transcribe after recording. Transcription runs on your device.
- Download your recording as MP3 or WAV, or copy the transcript. If you are signed in, use "Save to History" to keep it in your dashboard.
If the microphone does not work, check browser permissions (address bar or site settings) and ensure no other app is using the mic.
Voice Generator (text-to-speech)
Convert text into spoken audio. Steps:
- Paste or type your script in the text area. You can bring in text from the Script Editor or from your clipboard.
- Choose a voice and speed from the options.
- Click "Generate Audio." The synthesis runs in your browser. First load may take a moment while the model initializes.
- Play the result and use the download button to save the audio file (e.g. WAV) to your device.
Script Editor
Improve grammar and style of your text. Paste your script, then use the improve or polish actions. The editor is optimized for natural speech (e.g. scripts and voiceover text). Processing happens in your browser.
Text Summarizer
Paste long text (articles, meeting notes, transcripts) and choose format (e.g. paragraph or bullets) and length (short, medium, detailed). The summarizer runs locally and returns a concise summary. Best for meetings, lectures, and long articles.
VocEmail and other specialized tools
VocEmail turns voice notes or rough transcripts into professional email drafts. Other tools (Podcast Show Notes, Action Items, Meeting Minutes, Lecture Optimizer, Social Hooks, Clinical Note Architect, etc.) follow the same pattern: paste or bring in your text, adjust options, and generate. All run in the browser. Open the Tools page to see the full list and open any tool.
Saving work and accounts
Without an account, your work exists only in the current browser session. To keep recordings, transcripts, or generated text across devices and sessions, create a free account and use "Save to History" (or the equivalent) in each tool. Saved items appear in your Dashboard. You can also export files (download audio, copy text) at any time.
For more on privacy and data handling, see our Privacy Policy. For common questions, visit the Help Center.