Record. Protect.
Share safely.
The podcast studio that works without internet, without sending your data anywhere, and helps obscure your source's identity.
Coming soon to Google Play
"I interview whistleblowers. Their voices are pitch-shifted before the recording leaves my device. Everything runs offline — no data sent to any server."
— Daniel, 35, Central America. Investigative journalist.Voice Anonymization
Pitch-shift voice modification to help reduce speaker recognizability. Adjustable strength. Designed for source protection.
Fully Offline
AI noise reduction, voice anonymization, transcription — all on-device after one-time model download. No internet required. No data leaves your phone.
No Metadata
Exported audio files contain no location, device info, or timestamps in the file data. UUID filenames prevent timing analysis.
Field Recording
48kHz lossless recording with real-time monitoring, pause/resume, Bluetooth warning, and storage checks.
AI Noise Reduction
DeepFilterNet3 removes background noise on-device. One tap to studio quality — no cloud, no cost.
Waveform Editor
Cut, trim, undo/redo with pinch-to-zoom. Chapter markers for show notes. Export clips for social media.
On-Device Transcription
whisper.cpp running locally on your phone. Or bring your own OpenAI API key. Copy or share transcription text directly.
Double-Ender Recording
Each participant records locally. Auto-align by timestamp, mix with smart silence compression — no call quality compromise.
Clip & Share
Select a highlight, export as a clip, and share directly via WhatsApp, Telegram, or any app. Perfect for social media.
"I record interviews at the local market, remove the noise with one tap, cut the best parts, and share the episode — all from my phone. No PC needed."
— Priya, 32, Mumbai. Independent podcaster, publishes biweekly.A note on privacy
Kiku Studio is free, with no ads and no tracking. Voice anonymization uses pitch shifting, which helps reduce speaker recognizability but cannot guarantee complete anonymity — speech patterns, content, and environmental sounds may still allow identification by a determined adversary. For maximum protection, combine with device encryption and a strong lock screen.
Technical Details
- 48kHz / 16-bit mono PCM recording
- DeepFilterNet3 noise reduction (Rust/JNI, on-device)
- TD-PSOLA voice anonymization (pure Kotlin, no dependencies)
- whisper.cpp on-device transcription (C/JNI, ~75MB model download)
- EBU R128 loudness normalization (-16 LUFS)
- Energy-based Voice Activity Detection for silence trimming
- Non-destructive waveform editing with 20-level undo
- Chapter markers with timestamp export
- UUID filenames, no metadata in exports, FLAG_SECURE option
- Requires Android 8.0+, recommended Android 10+ with 4GB+ RAM
Listen with Kiku Player
Smart Speed, AI summaries, and real-time transcripts for any audio.
Learn about Kiku Player →