Back to deals
On-Device AI: Private Agent

Utilities

Free

On-Device AI: Private Agent

by Zijie Deng

4.4 (159) v1.44.0 1 729 MB Universal 17+

Description

Private AI chat, voice cloning, meeting transcription, web research, and agent workflows on your Mac, iPad, and iPhone. Run models locally or connect to cloud providers when you choose.

TALK, TRANSCRIBE, AND LISTEN
Turn voice into useful text, then turn answers back into speech when you want to listen instead of read.

• Three new TTS engines: Qwen 3 TTS, CostVoice, and VibeVoice
• Clone your voice and use it across multiple TTS engines
• Record meetings, lectures, interviews, and voice notes
• Import audio files and generate fresh transcripts
• Create natural speech with Apple voices, Kokoro, PocketTTS, or the new engines
• Export transcripts as text, subtitles, or markdown

Perfect for: meetings, lectures, research interviews, quick voice notes, and hands-free review.

AI WORKFLOWS YOU CAN SHAPE AS YOU GO
Real work rarely follows a straight line. Fork, copy, or export any conversation thread from the exact message that matters.

• Fork conversations mid-chat without losing the original
• Copy or export full threads for sharing or keeping records
• Long conversations stay fast — less lag as history grows
• Build subagents for research, writing, coding, analysis, or planning
• Run specialist agents on complex work and bring their results back into one conversation
• Switch models and tools as your task changes

PRIVATE BY DEFAULT
Your conversations, documents, transcripts, and local model runs stay on your device unless you choose to use an online provider.

• Run local models such as Llama, Gemma, Phi, Qwen, DeepSeek, and more
• Choose where models are stored — internal storage, external drive, or a shared volume
• Analyze sensitive documents offline
• Build searchable knowledge libraries for projects and research
• Connect to cloud providers only when you need them

WEB RESEARCH AND ACTIVE TOOLS
Give your AI access to current information when a task needs more than the model already knows.

• Search the web from inside a conversation
• Let the browser agent open pages, read sites, and gather information
• Pause browser work, adjust the task, or take over manually
• Use tools for calculations, network requests, document analysis, and research workflows

APPLE WORKFLOW INTEGRATION
On-Device AI is built for Apple devices, with shortcuts and workflows that fit into the way you already work.

• Mac menu bar access and customizable hotkeys
• Siri and Apple Shortcuts support
• iPhone, iPad, Mac, and Vision Pro support
• Hardware acceleration workflows across Apple devices
• Voice typing and dictation tools on Mac

Version 1.44.0 adds multi-language support, better STT and TTS voice enhancement, smoother chat performance, clearer TTS generation UI, improved speech model downloads, and bug fixes.

EULA: https://www.apple.com/legal/internet-services/itunes/dev/stdeula/

What's new (v1.44.0)

Version 1.44.0 makes On-Device AI easier to use globally and smoother for everyday voice work.

• Multi-language support: The app now has broader language coverage. Please send suggestions or wording feedback to developer@ondevice-ai.app.
• Better STT and TTS voice enhancement: Speech-to-text and text-to-speech workflows are more reliable, with improved handling around voice enhancement.
• Faster chat UI: Conversation screens have been optimized for smoother scrolling and more responsive long chats.
• Clearer TTS generation UI: Text-to-speech generation now gives a better visual flow while audio is being prepared.
• Improved model downloads: TTS and STT model downloads are easier to follow and more stable.
• Bug fixes and polish: Includes smaller fixes across voice, chat, and model workflows.