VoxMorph

VoxMorph is on-device voice cloning and text-to-speech for iPhone and iPad. Voice samples and text never leave your device. After the one-time model download, everything works offline , flights, secure environments, anywhere. ON-DEVICE, NOT CLOUD Most voice cloning services are token-based: every clone and every minute of audio counts against a monthly credit limit, and your recordings are stored on someone else's servers. VoxMorph is different: 100% on-device inference using Apple Silicon. Your voice samples never leave your iPhone, iPad, or Mac. No accounts, no third-party servers, no credit caps, no privacy compromise. Scale your eBook and audiobook production by running VoxMorph on multiple Apple devices with the same Pro license: clone on a Mac, polish on iPad, play back on iPhone. No cloud queues. No shared usage limits. NEW IN 1.1.1 - Soundboard , every project has a configurable grid (2x2 up to 8x8) of one-tap audition triggers. Bind cells to specific takes for live narration, podcast intros, performances. Great for projects exported from a Mac and played on iPhone with the same Pro license. - Project Bundles (.vxm) , share an entire project (audio, voices, soundboard, settings) as a single file. Clone voices on a Mac, then load the project on your iPhone to play back as a soundboard. - In-App Review , VoxMorph asks if you're enjoying it after a few successful projects. Or anytime from Settings , Rate VoxMorph. PRODUCING EBOOKS & AUDIOBOOKS VoxMorph is built for long-form production. Bulk-import a manuscript, organize sections, generate multiple takes per section (auditions) with varied speed and temperature, then pick the best read. Use different voices for different characters via per-section overrides. Export the whole audiobook with adjustable crossfade and loudness normalization. DEVICE TIERS - Voice Cloning requires a device with 10 GB or more RAM , iPhone 17 Pro, iPad Pro M4 16 GB, or any Apple Silicon Mac. - 8 GB devices (iPhone 15 Pro, 16, 16 Pro, 8 GB iPad models) run in TTS-only mode: full Voice Design (describe a voice in natural language) and Custom Voice (preset speakers with strong accent control) , but no recording-based cloning. - Sub-8 GB devices are not supported. KEY CAPABILITIES - On-device voice cloning powered by Qwen3-TTS , 3-30s sample, 24 kHz output - 10 synthesis languages: English, Mandarin Chinese (Beijing + Sichuan accents), Japanese, Korean, German, French, Russian, Portuguese, Spanish, Italian - Three voice variants: Voice Cloning (from a recording), Custom Voice (preset speakers + accents), Voice Design (described in natural language) - Live Mode , type a sentence, hear it in seconds - Transcribe , convert spoken audio to text on-device for editing - Project folders with shared voice/model/parameter defaults - Search across all sections and projects - Built-in audiobook export with crossfade + loudness normalization PRIVACY No analytics SDKs. No tracking. No third-party data brokers. Voice samples and text stay on your device. The only network requests are the initial model download from our content server, manifest checks, and subscription validation via Apple's StoreKit. SUBSCRIPTION Try free for 7 days. Monthly $2.99, Yearly $24.99 (save 30%), or one-time Lifetime $99.99. All tiers unlock identical functionality. https://www.apple.com/legal/internet-services/itunes/dev/stdeula/