High-accuracy AI transcription with speaker detection, AI summaries, and instant exports — all running locally on your machine. No cloud, no subscriptions, no data leaks. Ever.
Weaver was built for professionals who handle sensitive data — journalists, lawyers, doctors, researchers. Unlike cloud-based AI tools that upload your recordings to external servers, Weaver processes every second of audio locally on your machine using offline Whisper models and a local LLM. Your audio files and transcripts never leave your computer.
Whisper AI + Phi-3 LLM run entirely on your CPU/GPU. Nothing touches the internet after the one-time model download. Safe for legal, medical, and corporate work.
Automatically detects and labels different speakers using pause analysis. Optional AI refinement assigns descriptive names like "Dr. Chen (Guest)" — perfect for interviews and meetings.
Generate executive summaries, action items, and bullet points from any transcript — all powered by a local offline LLM. No cloud APIs, no token costs.
Beautiful modern UI with dark mode by default. Toggle to light mode anytime. Rounded buttons, smooth layout, and clean typography throughout.
Save transcripts as Word (DOCX), plain text (TXT), or SRT subtitles — with accurate timestamps and speaker labels preserved in every format.
Drag-and-drop MP4, MP3, MOV, M4A, or WAV files. Weaver auto-converts any media to high-fidelity 16kHz WAV for optimal transcription accuracy.
Ctrl+O to open, Ctrl+T to transcribe, Ctrl+R to record. Export with Ctrl+Shift+S/W/E for Word/TXT/SRT. Power-user workflow built in.
Capture any audio playing through your speakers — Zoom calls, podcasts, webinars. Auto silence detection prevents empty recordings.
Pay once. Own forever. No monthly "AI tax." All v2.x updates included free for lifetime license holders.
Buying a Lifetime License today locks in your price and guarantees you receive all upcoming features for free.
Drop a folder of audio or video files and let Weaver process them all sequentially in the background — ideal for large archives.
Adding highly optimized models for German, French, Spanish, Mandarin, Portuguese, Arabic, and more.
Direct microphone input alongside system audio capture — perfect for in-person meetings and dictation.
Bringing Weaver's offline AI engine natively to macOS (Apple Silicon) and Linux distributions.
✅ Already shipped in v2.0: Speaker Diarization, AI Speaker Refinement, Dark/Light Themes, TXT/SRT Export, Keyboard Shortcuts, Drag & Drop, Real-time Progress Bar, and Settings Persistence.
Lifetime License — v2.0
One-time payment. All v2.x updates included.
Secure checkout via Lemon Squeezy
*Note on Offline Use: An internet connection is required on first launch for license activation and a one-time download of the AI models (~2-3 GB). After setup, Weaver runs 100% offline.