Voice Rack: AI Speech to Text
Local AI for recorders & memos
Only for Mac
Free · In‑App Purchases
Mac
Why 4.8 GB? Inside is a 4-billion-parameter language model (3.5 GB) plus speech and search models, all running on your Mac itself: offline, nothing uploaded.
Turn a folder of recordings into a library you can ask.
Voice recorders, phone memos, meeting captures — scattered everywhere, named nothing. Point this app at a folder and minutes later every recording has a summary and tags; you can filter, search, and ask.
Fully offline is the architecture, not a slogan.
The latest on-device AI models ship inside the app. No account, no subscription activation, no network permission at all. Your recordings never leave your machine.
Why this is not another transcription app:
• Ask your whole library — "Which meeting discussed the budget? What was decided?" Answers cite the source file and timestamp; click to jump to that second. Ask in one language, match recordings in another.
• Chat with a single recording — follow-up questions with context, every answer time-stamped for instant verification.
• A library in seconds — organized by metadata first; transcription runs in the background, and if you quit midway it picks up where it left off when you choose Resume.
• A term book that learns — names and jargon get transcribed wrong once; fix them once, corrected forever. It can even learn terms from your own documents.
• Honest by design — uncertain passages are flagged for a three-second listen instead of dressed up; when the transcript has no answer, it says so.
Free is the full app for one recording. The large on-device model — not a small "lite" one — ships inside and runs from the first launch with nothing to download. Speaker identification, a term book that learns, summaries, seven export formats: all free, per file, forever. Keyword search and tag filtering across the whole library are free too. There is no cloud tier and nothing to sign in to.
Pro does the same for your whole library. One purchase, no subscription: ask questions across every recording with cited sources; transcribe and summarize the whole library in one go; apply term corrections back across the whole library; gather every highlight from every file in one place; merge duplicate tags across the whole library; compress the whole library; export the current list in one step. 14-day free trial, then a one-time purchase.
Speaks your languages. Transcribes 90+; measured end to end: English, Chinese (Mandarin & Cantonese), Spanish, French, German, Dutch, Italian, Japanese, Korean, Russian. Ask in your language — answers come back in it, even when the recording speaks another. Summaries and tags are written in the app's display language; transcripts always stay in the recording's own language.
Everything: batch import · watch folders · sentence-level timestamps · multilingual transcription · local translation · seven export formats (SRT/VTT subtitles, Markdown, CSV, HTML, TXT, JSON) · smart tags · full-text search · multi-turn Q&A · cross-library retrieval
Requirements: Apple Silicon Mac (M1 or later), 16 GB RAM recommended, 8 GB minimum. About 5 GB installed — the models live inside the app; that is what offline costs, and what it guarantees.
Ratings & Reviews
- This app hasn’t received enough ratings or reviews to display an overview.
• Copy to the clipboard: the Export menu now has "Copy This Page (with timestamps)" and "Copy This Page (plain text)". They copy whichever tab you are viewing (Original, Cleaned, Summary, Translation or To-dos), ready to paste into notes, a chat or an email.
The developer, Zhang Jie Jing, indicated that the app’s privacy practices may include handling of data as described below. For more information, see the developer’s privacy policy .
Data Not Collected
The developer does not collect any data from this app.
Accessibility
The developer has not yet indicated which accessibility features this app supports. Learn More
Information
- Seller
- Zhang Jie Jing
- Size
- 4.7 GB
- Category
- Productivity
- Compatibility
Requires macOS 26.0 or later and a Mac with Apple M1 chip or later.
- Mac
Requires macOS 26.0 or later and a Mac with Apple M1 chip or later.
- Mac
- Languages
English and 8 more
- English, Dutch, French, German, Japanese, Korean, Simplified Chinese, Spanish, Traditional Chinese
- Age Rating
4+
- 4+
- In-App Purchases
Yes
- 14-day Trial $0.00
- Voice Rack Pro $49.99
- Copyright
- © 2026 Jiejing Zhang
