.TXT
UTF-8 · TEXT
Clean text for writing
Clean, readable plain text organized into natural paragraphs — ideal for drafting articles, meeting minutes, or reports.
Articles
Drafts
Docs
FREE DESKTOP APP / WINDOWS 10 & 11
Offline Speech to Text.
Free on Windows.
Accurate transcription of audio, meetings
and video into clean text.
Slysh! is a free, 100% offline alternative to MacWhisper for Windows 10 and 11. Powered by local OpenAI Whisper and NVIDIA NeMo, it transcribes audio to text, records meetings without bots, separates speakers, and generates millisecond-accurate SRT subtitles with zero subscription fees.
Local models · Free cloud tiers · Windows 10 & 11
Click any phrase in the transcript
to hear it directly in audio playback.
ACCURATE MULTILINGUAL WHISPERBOT-FREE MEETINGS IN ZOOM, MEET & TEAMS100% PRIVACY — AUDIO STAYS ON YOUR PCVOICE TYPING ON WINDOWS PCOFFLINE TRANSCRIPTION WITHOUT INTERNETSPEAKER DIARIZATION (WHO SAID WHAT)YOUTUBE CHANNELS & PLAYLISTS TO SUBTITLESFREE CLOUD AS AN OPTION (GROQ & GLADIA)NO SUBSCRIPTION FEESEXPORT TO SRT · VTT · TXT · MD
01 / YOUR USE CASES
From a single quote to a full archive
Less relistening.
More productive work.
Audio and video transcription workflows:
from quick soundbites to meeting archives.
01
Interviews & podcasts
Find the exact quote, verify phrasing against audio playback, and separate speaker turns.
Explore text editor 02Work meetings
Capture remote colleagues and your own microphone without inviting awkward bots. Revisit key decisions anytime.
Explore meeting recorder 03Lectures & research
Extract clean text or existing YouTube subtitles. Search topics and save reference material to your knowledge base.
Explore YouTube tool 04Video & subtitles
Transcribe single clips or entire folders. Polish copy and export timestamped subtitles ready for video editing.
Explore export formats02 / CAPABILITIESFor those who work with words
You speak.
Slysh! captures what matters.
From audio and video transcription to ready-to-use subtitles and PC voice typing.
All inside one unified workspace.
ACCURACY & PRIVACY
Local Whisper speech to text
Accurate multilingual speech recognition on your PC: handles technical terms, accents, and fast dialogue without hallucinations.
A free Windows alternative to MacWhisper: accurate offline speech recognition powered by OpenAI Whisper on your GPU, handling technical jargon, multiple speakers, and diverse accents without cloud dependencies.
MEETINGS & CALLS
Meeting recording without bots
Direct capture of Windows system audio and your microphone in Zoom, Meet, and Teams without inviting third-party bots.
Slysh! records loopback audio and your microphone directly, never disturbing participants or requiring host permissions.
YOUTUBE CONTENT
YouTube: channels and playlists
Batch extraction of existing subtitles from videos and playlists straight to SRT and VTT without downloading video.
No need to download gigabytes of video: extract clean author or automatic captions with millisecond timestamps.
ARCHITECTURE
Locally on PC or in the cloud
100% private offline on GPU, or plug in generous free quotas from Groq, Gladia, and Deepgram without installing PyTorch.
Dedicated GPU? Work entirely offline. Lightweight laptop? Connect free cloud APIs with zero system strain.
TASK QUEUE
Batch processing for multi-file queues
Add dozens of audio files or folders: sequential background processing without interface stutter or lag.
Configure models, languages, and auto-export: Slysh! processes the entire queue smoothly in the background.
STORAGE & ARCHIVE
Local archive in SQLite database
Built-in synchronized editor, audio player, and local database with instant full-text search across your library.
All transcripts stored securely on your disk in SQLite: easy playback scrubbing, in-place corrections, and multi-format export.
03 / INSIDE SLYSH!From recording to polished copy
Every word
in its right place.
Timestamped segments, in-place editing, and clean export.
See how it works inside the app.
Slysh! — Transcript Editor
04 / EDITOR & PLAYER
Synchronized waveform & precise edits
Find any quote —
without listening twice.
A synchronized desktop audio to text editor. Transcripts stay in sync with audio playback. Click any word to jump playback, correct terms in place, and save changes straight to your local archive.
Player and transcript side-by-side — no need to toggle between multiple applications.
Full-text search across transcript
Locate key discussion topics, participant names, or specific phrases in fractions of a second.
Jump to audio on click
Click any sentence to hear it instantly. The active segment highlights in real time during playback.
In-place text & timestamp editing
Correct technical terminology and wording while keeping precise millisecond timestamps intact.
Progressive transcript reading
In supported modes, review and edit earlier segments before the entire long recording finishes processing.
05 / BOT-FREE MEETINGS
Record meetings.
Without unwanted bots.
Slysh! captures Windows system audio and your microphone directly via WASAPI loopback. Works with Zoom, Google Meet, Microsoft Teams, Slack, and any desktop call.
No intrusive meeting bot ever joins your call, makes participants uncomfortable, or requires host permissions. Complete confidentiality for sensitive interviews under strict NDA.
Speaker diarization never uploads voice biometrics online — all processing runs locally on your PC.
06 / DIARIZATION & SPEAKERS
Speaker diarization.
Know who said what.
Automatic neural voice separation by NVIDIA NeMo Sortformer. Split recordings into distinct speaker turns, label participant names, and export structured dialogue.
- Identify speakers after transcription or enable diarization automatically in batch processing rules.
- Rename a participant once — their name updates across all dialogue turns in the entire transcript.
- Retain speaker names across supported export formats, including Markdown (.md) and SRT subtitles.
NVIDIA NEMO
SORTFORMER
1-CLICK ATTRIBUTION
100% PRIVATE
NEMO SORTFORMER
Interactive dialogue
Offline on PC
00:12
Anna
Let’s start with the user interviews. What is the most critical question for participants?
00:19
Michael
How they process recordings today and where they lose the most time.
00:27
Sarah
I’ll draft the interview questions and gather all reference materials by Friday.
00:34
Anna
Sounds great. Once the interviews wrap up, we’ll jump straight to exact quotes.
07 / YOUTUBE CONTENT
Without downloading heavy video files
YouTube subtitles.
From single clips to full channels.
Extract subtitles and convert YouTube video audio to clean text without downloading gigabytes of video. Paste a link to a video, playlist, or entire channel queue.
01
Select source materials
Set publication date ranges and video count limits. Exclude short-form Shorts with a single toggle.
02
Preview the queue
See which videos will be fetched and check which titles already exist in your local library.
03
Build your knowledge library
Configure automatic export to Markdown or SRT for immediate reference, research, and content work.
OPENAI WHISPER & FASTER-WHISPER
NVIDIA CUDA ACCELERATION (GPU)
SPEAKER DIARIZATION NEMO SORTFORMER
THE MACWHISPER ALTERNATIVE FOR WINDOWS
FREE CLOUD OPTIONS: GROQ & GLADIA
LOCAL SQLITE DATABASE ARCHIVE
SYNCED AUDIO WAVEFORM & TIMESTAMPS
SUBTITLES ACCURATE TO MILLISECONDS
08 / BATCH PROCESSING
High-throughput queue without UI lag
Add a folder.
Process files in batch.
When handling dozens of recordings, there is no need to trigger each manually. Simply drop multiple files or an entire folder — Slysh! transcribes the queue in the background.
1
Collect audio & video
Add multiple MP3, WAV, M4A, MP4, MKV files or a directory. Duplicate prevention ensures you never re-transcribe completed files.
2
Configure settings
Set speaker diarization, desired export formats, and choose whether to save outputs alongside source files or in a target directory.
3
Export results
Jump between transcripts right in the editor and export the entire batch in one click without tedious manual steps.
09 / EXPORT & ARCHIVE
Ready for publishing & long-term storage
Export text and subtitles.
To the right formats and folders.
For articles. For team knowledge bases.
For your next video production.
.MD
MARKDOWN
Knowledge base assets
Structured Markdown with metadata and speaker tags — perfect for Obsidian, Notion, Logseq, and note-taking apps.
Notes
Sources
Obsidian
Notion
.SRT / .VTT
AUDIO TO SRT
Subtitles for video
Fast audio to SRT and VTT subtitle conversion with millisecond precision. Ready for import into Premiere Pro, DaVinci Resolve, or YouTube Studio.
Audio to SRT
Video Editing
Subtitles
Premiere / DaVinci
AUTOMATIC ROUTING
One export. Multiple destinations.
Save transcripts simultaneously to your project directory, Obsidian vault, and team shared drive. Open the file or its folder with a single click after export.
Ready transcript
1 click → 3 destinations
Project folder
SYNCED
Knowledge base (Obsidian)
SYNCED
Team folder
SYNCED
10 / PRIVACY & DATA CONTROL
Offline on your PC.
Or in seconds via cloud.
Dedicated GPU? Transcribe gigabytes of audio locally with zero data exposure.
Lightweight office laptop? Connect Groq or Gladia in 1 minute without installing heavy neural packages.
File ControlAudio is never uploaded to external servers without your explicit choice. Complete confidentiality for private discussions.
No Slysh! SubscriptionsZero monthly fees, per-minute charges, or vendor lock-in. Run free offline or plug in your own API keys.
Unified SQLite ArchiveAll transcripts are organized into a single local SQLite database on your disk with instant search.
WHISPER & NEMOCUDA / CPU100% OFFLINE$0 / MINUTE
Process speech entirely on your computer. Once the model is downloaded, no internet is required for transcription.
Internet:Not required (100% offline)
Requirements:CUDA (NVIDIA GPU) or CPU
Privacy:Strictly on your local disk
- Audio is processed locally on your machine
- No per-minute API fees or subscription costs
- Local stack requires available disk space and compute
Local AI Whisper and NeMo running offline on your PC
FREE CLOUD ONBOARDING
When you prefer zero PC compute strain
Free cloud tiers.
Ideal for lightweight laptops.
Free transcription in the cloud. Create an account with a supported provider, paste your personal API key, and transcribe speech without warming up your CPU or draining battery.
Groq
Free quota:
8 hrs / day
Cost:
$0
- Fast Whisper on LPUs
- 2,000 requests per day
- 0% CPU and GPU load
Gladia
Bonus credit:
+€50
Cost:
$0
- Speaker diarization
- Enough for dozens of meetings
- No credit card required
Deepgram
Free grant:
+$200
Cost:
$0
- Instant enterprise ASR
- Hundreds of hours of audio
- No billing info required
Complete freedom of choice: Slysh! runs 100% locally without internet on your PC, with cloud as a flexible complement whenever you want to keep your laptop cool. Slysh! is completely free: simply paste your free API key into app settings.
11 / THE PERSON BEHIND THE PROJECT
Crafted with care
Built for real tasks.
Growing with your support.
Creator of Slysh!
12 / FREQUENT QUESTIONS
What you should know before starting
Good questions.
Clear answers.
Everything you need to know about offline processing, privacy, supported formats, and system requirements.
Do I need an internet connection?
Not for local transcription once dependencies and model weights are installed on your computer. An internet connection is only needed for cloud providers (Groq, Gladia) and fetching YouTube subtitles.
Where are my audio recordings sent?
In local mode, all audio processing occurs directly on your PC and never touches external servers. In cloud mode, audio is sent strictly to the provider API you configure yourself.
What if I don't have a powerful GPU?
Use cloud-only mode with free API tiers (Groq offers generous daily free hours, Gladia provides starter credits). Slysh! can also run Whisper models on a standard CPU.
Can I install Slysh! on Mac or Linux?
Currently, Slysh! is specifically tailored for Windows 10 and Windows 11 (64-bit) with hardware acceleration for NVIDIA GPUs and direct Windows loopback audio capture.
Is there a MacWhisper equivalent for Windows?
Yes. Slysh! is designed as the offline Whisper transcription app for Windows 10 and 11. It provides the same seamless local experience as MacWhisper: synchronized playback, waveform navigation, speaker diarization, and multi-format subtitle export with native NVIDIA CUDA acceleration and zero subscriptions.
How does Slysh! differ from Telegram bots and paid online services?
Complete data privacy, zero subscriptions, and offline capability. Your recordings are protected by NDA, never stored on foreign servers, and there are no per-minute transcription limits on your PC.
Which export formats are supported?
Transcripts can be exported to clean TXT, structured Markdown (.md) with speaker tags and timestamps, and industry-standard subtitles (SRT and VTT).
Does YouTube batching work with any video?
The YouTube feature retrieves existing subtitles and captions without downloading heavy video files. For custom video files, use standard transcription.
How do I transcribe audio and video to text offline for free?
Install Slysh! on your Windows computer and select local Whisper models. The software transcribes audio (MP3, WAV, M4A) and video (MP4, MKV) 100% offline using your PC's hardware, without subscriptions or minute caps.
How does voice typing on PC in Windows work?
Slysh! features a 1-click dictaphone that uses state-of-the-art neural speech models. Speak into your microphone, and Slysh! accurately transcribes your speech with punctuation directly into the editor.
Can I use Slysh! as a free alternative to Descript or Otter.ai?
Yes. While Descript and Otter limit free accounts to 30–60 minutes per month and upload audio to cloud servers, Slysh! runs 100% locally on your PC. There are no monthly minute limits, no bots joining your calls, and complete confidentiality under NDA.
Is it truly 100% free?
Yes, the software itself is free and open-source with no fees or commissions. Local recognition runs on your PC hardware, and cloud modes connect to provider APIs using your personal keys and free tier quotas.
SPECIFICATIONS & COMPATIBILITY
WINDOWS 10 / 11 · 64-BIT
Technical parameters and formats
Audio input formats
Video input formats
Text & subtitle export
AI recognition models
Hardware requirements
Any modern CPU or NVIDIA GPU (CUDA) with 4–6 GB VRAM
Data privacy
100% local on PC: audio recordings and SQLite database are never uploaded online
FREE & OPEN SOURCE TOOL
Ready to try Slysh! in action?
Run it on your Windows computer. Transcribe meetings, audio files, and video with zero per-minute fees, bots, or limits.
Download for WindowsWindows 10 / 11 · Convenient installer