Local AI Transcription& Instant System Dictation
100% private, on-device Whisper speech-to-text for macOS & Windows. Dictate anywhere with global hotkeys, remove filler words automatically, and batch transcribe long recordings with zero latency.
BUILT FOR SPEED, PRECISION & ABSOLUTE PRIVACY
Designed from the ground up for professionals, creators, researchers, and developers who demand fast, private speech processing.
100% On-Device Privacy
Powered by Whisper.cpp running locally on your hardware. Audio recordings and transcripts never touch an external server or third-party cloud unless explicitly requested.
System-Wide Dictation
Trigger dictation anywhere with a customizable global hotkey (Option + Space / Ctrl + Alt + E). Speak naturally and instantly paste formatted text right into your cursor.
Local AI Text Cleanup
Integrated sidecar LLM (Qwen 1.5B) cleans up filler words ("um", "uh", "you know"), fixes grammatical stumbles, adds punctuation, and formats paragraphs cleanly in real-time.
Word-Level Timestamps
Click any word chip in the transcript to jump playback to that exact second in the audio file. Precise timecodes simplify editing, captioning, and quote verification.
Speaker Diarization
Automatically identify and label different speakers using on-device small.en-tdrz models in multi-person podcast interviews, team meetings, and audio recordings.
Modal Cloud GPU (~95x Realtime)
Need ultra-fast processing for huge multi-hour audio files? Opt-in to serverless Modal L4 GPU cloud acceleration backed by Whisper Large v3 whenever you choose.
ACTUAL APP SCREENSHOTS
Take a look inside the native EchoShift desktop workspace.
Word-Level Transcript & Playback
Interactive word-by-word timestamp chips, search filter, and instant audio playback player.
Multi-Format Exporting
Export directly to Plain Text (.txt), Markdown (.md), Subtitles (.srt), WebVTT (.vtt), Word-level JSON, or Cleaned Prose.
Redeem Code Activation & Referrals
Redeem your 16-character access code, view your 3 shareable invite codes, and manage cloud credits.
On-Device AI Model Manager
Choose between Whisper Base (Fast), Small (Balanced), Large (Accurate), Qwen Cleanup, and Speaker Detection models.
Dictation & Cleanup Preferences
Set transcription language, toggle Polished rewrite mode, or enforce verbatim dictation for code editors.
Cloud GPU Telemetry Dashboard
Real-time telemetry showing Modal L4 GPU execution speed (~95x realtime speedup) and balance.
FREE DOWNLOAD WITH REDEEM CODE UNLOCK
EchoShift is completely free to download for anyone on macOS and Windows. Activation requires a 16-character Redeem Code (e.g. ES-ABCD-1234-EFGH or ACCESS-...). Once unlocked, every user gets 3 invite codes to share with colleagues.
Download App
Grab the free installer for macOS (.dmg) or Windows (.msi) directly below.
Launch & Enter Email
Open EchoShift and enter your email address to initiate activation.
Enter Redeem Code
Paste your ES-... or ACCESS-... redeem code to receive your magic sign-in link.
Unlock & Share
Unlock full local transcription immediately and receive 3 invite codes of your own.
ES-ABCD-1234-EFGHTECHNICAL SPECIFICATIONS
Hardware requirements and compatibility details for EchoShift.
| macOS Requirements | macOS 12.0 (Monterey) or later. Native Swift support for Apple Silicon (M1/M2/M3/M4) & Intel. |
|---|---|
| Windows Requirements | Windows 10 / Windows 11 (64-bit). Built on Tauri 2.x with WASAPI low-latency audio capture. |
| AI Models Supported | Whisper Base, Small, Large v3 (on-device) + local sidecar Qwen 1.5B LLM for transcript polishing. |
| Supported Inputs | MP3, WAV, M4A, AAC, FLAC, OGG, MP4, MOV, AVI, MKV, WebM, & Live Microphone Input. |
| Export Formats | Plain Text (.txt), SubRip Subtitles (.srt), WebVTT (.vtt), Word-level JSON (.json), Markdown (.md), Cleaned Prose. |
FREQUENTLY ASKED QUESTIONS
READY TO EXPERIENCE PRIVATE AI DICTATION?
Download EchoShift for free today. Have your redeem code ready to activate your full license on macOS or Windows.