Free AI audio and video transcription with timestamps, captions, and subtitle translation.
Zen Whisper
Summary
Local-first Mac dictation that types into any app, plus file and link transcription across 110 spoken languages on Apple Silicon.
Screenshots
Description
Zen Whisper is a dictation and transcription app for macOS that keeps core speech recognition on the machine instead of streaming your voice to a server. Hold a keyboard shortcut, speak, and the text lands at the active cursor in whatever app you are already in — no separate capture window to copy out of.
Dictation that follows the cursor
Because it inserts at the typing caret, Zen Whisper works anywhere on the Mac you can normally type or paste: a mail composer, a code editor, a browser form, a chat box. There is no per-app integration to install.
Transcription beyond the microphone
The same engine transcribes uploaded audio and video files and supported public links. Finished jobs go into a searchable history alongside voice memos, snippets and a dictionary memory that learns the names and terms you use, so a transcript you made last month is still findable.
Local-first, and multilingual by design
Speech packs are downloaded and kept on-device, and the app is tuned for low latency on Apple Silicon — it skips heavy idle prewarming on lower-memory Macs and unloads native models after a quiet period rather than holding RAM indefinitely. The app exposes 110 spoken language options, with unusually strong coverage of Indian and regional languages including Hindi and Hinglish alongside English variants. Any network-backed extras are opt-in and documented separately in the privacy policy.
Requirements and pricing
Zen Whisper is a native Apple Silicon app requiring macOS 14 or later; there is no iPhone or iPad version. It starts with a 14-day free trial and then remains usable at 3,000 words per week on the free tier. Pro is sold on Gumroad as seat packs for 1, 2, 5 or 10 Macs, each a lifetime licence for that bundle with one year of updates included and discounted renewal afterwards; Pro adds larger models, media workflows, rewrite tools and translation where available.
It is made by ZENPRODUCTS Technologies Private Limited, whose other apps follow the same on-device, privacy-first pattern.
Reviews
Similar App Suggestions
Voice typing that rewrites rough speech to fit the app you are in — Slack, Gmail, Cursor or a notes field — across 100+ languages.
Alibaba's AI music model: describe a mood, story or style in one line and get a finished song with lyrics, vocals and arrangement.
Speech-to-text service for meetings, interviews and video in 100+ languages, with an AI chat that answers questions against your transcripts.
AI note taker that transcribes meetings, interviews and recordings in 58 languages, then summarises them and pulls out action items.