Your Voice, Your Data, Your Device.
Privacy-first AI transcription, OCR & document processing for macOS, iOS, and Chrome. Record, transcribe, extract text, and enhance—all on-device. No uploads. No tracking. No internet required.
4 transcription engines. 8 AI providers. All run on Apple Silicon. Choose local-only for complete privacy, or add cloud APIs for extra accuracy.
One Workflow. Three Platforms.
Start recording on Mac, continue on iPhone, or capture meetings with Chrome. All transcripts sync locally via iCloud (optional).
Full-Power Desktop
- • 3-column layout (sidebar + editor + inspector)
- • Full keyboard shortcuts, menu bar commands
- • MCP server for Claude Desktop/Code integration
- • Meeting server for Chrome extension (localhost:8085)
- • Drag-and-drop files, window management
- • Batch processing, directory monitoring
Capture On-the-Go
- • iPad: Split view, multitasking, drag-and-drop, trackpad support
- • iPhone: Tab bar navigation, floating record button, swipe actions
- • Haptic feedback, lock screen controls, mini player
- • Quick voice notes with widget
- • Offline transcription anywhere
- • AirPods/external mics, screen capture
Meeting Companion
- • Google Meet + Microsoft Teams (teams.live.com, teams.microsoft.com)
- • Quality settings: Low/Medium/High, audio-only option
- • Meeting metadata: title, participants, timestamp
- • Auto-stop when leaving, auto-delete after processing
- • Connection status indicator (green=connected, red=offline)
- • Auto-upload to macOS app (localhost:8085)
Record → Transcribe → Enhance
Three steps to perfect transcripts. No technical skills required.
Hit Record
Use built-in mic, AirPods, or professional audio interface. Real-time visualization shows you're capturing clearly.
Choose Your Engine
4 transcription engines: WhisperKit (99 langs, offline), FluidAudio (50× faster), OpenAI Whisper API (cloud accuracy), or Apple Speech (built-in). Pick based on speed, accuracy, or privacy needs.
Enhance with AI
8 AI engines to choose from: Local (MLX, Apple Intelligence, Claude Code CLI, Gemini CLI, Codex CLI, Copilot CLI) or Cloud (Claude API, OpenAI API). Fix grammar, generate summaries, translate, extract key points. Fully custom prompts supported.
Everything You Need, Nothing You Don't
Choose Your Transcription Engine
WhisperKit (local, 99 langs, Tiny~40MB to Large-v3-Turbo~3GB models), FluidAudio (local, 55 langs, 50× faster), OpenAI Whisper API (cloud, highest accuracy), Apple Speech Analyzer (built-in, 45+ langs). Word-level timestamps for precise alignment.
Choose Your AI Engine
8 LLM providers: Local (MLX, Apple Intelligence) or Cloud (Claude API, OpenAI API) + CLI engines (Claude Code CLI, Gemini CLI, Codex CLI, Copilot CLI). AI Intelligence Pipeline extracts Action Items, Decisions, Questions, Topics. Output templates: Meeting Minutes, Lecture Notes, Interview/Podcast Highlights. Smart chunking adapts to context windows (4K-1M tokens).
Smart Speaker Diarization
Auto-detects speakers with LLM-based correction. Perfect accuracy for interviews, meetings, panels. Edit speaker names inline.
Visual Intelligence & Documents
Video frame extraction + OCR on frames. Attach PDF, DOCX, PPTX, XLSX, images as reference materials—all OCR'd and indexed for semantic search across both transcript AND document content.
Process Any Document
PDF, DOCX, PPTX, XLSX support. Extract content, enhance with AI, export to TXT, MD, CSV, JSON, or ZIP. Batch process entire folders.
Live Transcription
Two modes: Normal (Accurate, 8-10s chunks) and Fast (Low-Latency, 3-5s chunks). Auto-save partial transcripts every 30s for crash recovery. Real-time word count, segment counter, latency display. Auto-scroll to latest transcript.
Organize & Find Anything
7 export formats: SRT, VTT, Markdown, PDF, JSON, CSV, TXT. Export options: include/exclude timestamps, speaker labels, translations. Live preview before export. Batch export multiple recordings. Custom folders, tags, favorites, archive with hybrid search.
Custom AI Prompts
Go beyond grammar/summary/keypoints—create completely custom prompts for your use cases: meeting action items, legal brief extraction, medical notes, code review summaries. Save templates for reuse.
Translation
Apple Translation framework for bilingual display mode. Auto-detect source language, on-demand language packs. Translate transcripts without cloud services.
Custom Dictionary & Auto-Correction
User-defined word→replacement pairs. Built-in packs: Medical, Legal, Technology, Business. Regex support, auto-apply after transcription. Perfect for domain-specific terminology.
Audio Editing
Trim audio with waveform handles, split into clips, auto-skip silent segments. Non-destructive editing. Export clips as m4a/wav/mp3.
Advanced Segment Editor
Merge/split segments, undo/redo stack, find & replace, keyboard navigation (Mac), flag segments for review. Professional-grade transcript editing.
Playback Controls
Speed control 0.5x-2x, skip forward/backward (5s/15s/30s configurable). Tap word in transcript to seek to exact timestamp. Waveform scrubbing for visual navigation.
Accessibility
Full VoiceOver support, Dynamic Type, WCAG 2.1 AA contrast. Keyboard navigation on Mac, Reduce Motion support. Built for everyone.
Never Miss a Meeting Moment Again
MinuteAI Chrome Extension works with Google Meet and Microsoft Teams (teams.live.com, teams.microsoft.com). Record, transcribe, and review—all while staying present in the conversation.
- Auto-record on meeting join (optional)
- Live transcript in side panel + floating widget overlay
- Manual start/stop controls
- Quality presets: Low (bandwidth saver) | Medium | High
- Audio-only mode (no video capture)
- Meeting metadata collection (title, participants, timestamp)
- Real-time upload to MinuteAI macOS app (localhost:8085)
- Auto-stop when leaving meeting, auto-delete after processing
- Connection status indicator (green=connected, red=offline)
- Supports 10 languages: EN, JA, VI, KO, ZH, FR, DE, ES, PT, TH
- Free forever (no usage limits)
AI-Powered Enhancement & Developer Tools
Choose your AI engine for enhancement, or build custom workflows with our MCP server and APIs.
Model Context Protocol Server
macOS onlyFull MCP implementation with 11 tools. Use MinuteAI directly from Claude Desktop, Claude Code, or Cursor. JSON-RPC 2.0 protocol.
-
list_recordings,get_transcript,search_recordings -
get_speakers,enhance,retranscribe - + 5 more tools for full automation
Local CLI AI Engines
No API Keys RequiredUse your local AI CLIs as enhancement engines in-app. MinuteAI connects automatically—no API keys needed. Get AI summaries, translations, grammar fixes, and custom prompts for free if you already have these tools installed.
- Claude Code CLI - Use as AI engine, zero config
- Gemini CLI - 1M context window, local
- Codex CLI + Copilot CLI - Coding-focused AI
- Great for developers: free AI enhancement via tools you already use
Smart Search & Related Recordings
Find anything across your entire library. MinuteAI automatically discovers related and similar recordings using semantic understanding — so you can connect meetings, lectures, and notes by meaning, not just keywords.
- Full-text search across all transcripts & documents
- Semantic similarity finds related recordings automatically
- "Show Related" surfaces connected recordings by topic & context
Example workflow: Record meeting → Auto-transcribe with FluidAudio → Enhance with Claude Code CLI (no API key) → Generate action items → Export to Notion via MCP server
Download & Start BuildingPrivacy-First, You Choose
Local engines (WhisperKit, FluidAudio, MLX, Apple Intelligence) keep data on-device. Cloud options (OpenAI API, Anthropic Claude) available when you need higher accuracy. You decide.
Local-First by Default
Default engines (WhisperKit, FluidAudio, Apple Speech, MLX, Apple Intelligence) process on-device. Cloud APIs (OpenAI, Claude) are optional for users wanting higher accuracy. Your audio never touches our servers.
No Analytics, No Ads
We don't track what you record, who you talk to, or what you say. No third-party analytics. No ad networks.
You Control Storage
All recordings and transcripts save to your device. Optional iCloud sync uses end-to-end encryption managed by Apple, not us.
Transparent & Auditable
Our privacy policy is written in plain language. No legal jargon. No hidden clauses. Read it in 2 minutes.
Built for How You Work
Remote Workers
"I record every standup and 1-on-1. MinuteAI gives me transcripts instantly so I can focus on the conversation, not note-taking."
Stay present in meetings
Journalists
"Interviewing sources for hours used to mean days of transcription. Now I get word-perfect transcripts in minutes."
Faster turnaround
Students
"Lectures move fast. I record everything and review transcripts when I study. Speaker diarization even labels different professors."
Better retention
Content Creators
"I batch-process all my podcast episodes overnight. Wake up to 20 transcripts ready for blog posts and show notes."
Repurpose content
Legal Professionals
"Client confidentiality is non-negotiable. On-device processing means sensitive discussions never leave my MacBook."
Compliance confidence
Start Free. Upgrade When You Need More.
No credit card required. No trials that expire. Free forever with generous limits.
Free
- Unlimited recordings (up to 10 minutes each)
- On-device transcription (WhisperKit, FluidAudio, Apple Speech)
- Basic AI enhancement
- Export to TXT, Markdown
- Chrome Extension included
- Speaker diarization (up to 3 speakers)
- iCloud sync (optional)
Limits: Batch processing (5 files), Enhancement (10/month)
Download FreePro
Everything in Free, plus:
- Unlimited batch processing
- Unlimited AI enhancement
- Advanced summaries & action items
- Export to PDF with formatting
- Priority support
- Speaker diarization (unlimited)
- Custom enhancement prompts
🎁 7-day free trial (no credit card)
Upgrade to ProBuilt for People Who Handle Sensitive Audio
If your recordings contain confidential information, local processing matters.
Journalists
Protect source confidentiality. Interview audio stays on your device.
Lawyers
Keep attorney-client communications off third-party servers.
Researchers
Fieldwork transcription that works offline. IRB-friendly by design.
Students
Record lectures and get searchable transcripts. Free for recordings under 10 min.
Business
NDA-safe meeting transcription. No third-party data processing.
Compare
See how MinuteAI compares to Otter.ai and Descript — features, pricing, privacy.
Frequently Asked Questions
No. All transcription and AI processing happen on-device using your choice of engines — WhisperKit, FluidAudio, or Apple Speech for transcription; MLX, Apple Intelligence, or local AI CLIs for enhancement. Cloud APIs (OpenAI, Claude) are optional.
WhisperKit supports 99+ languages for transcription. The Chrome Extension UI is available in 10 languages: English, Japanese, Vietnamese, Korean, Chinese (Simplified), French, German, Spanish, Portuguese (Brazil), and Thai.
macOS: macOS 14 (Sonoma) or later with Apple Silicon (M1, M2, M3, M4) or Intel with discrete GPU. iOS/iPadOS: iOS/iPadOS 17+ with A12 Bionic or newer (iPhone XS/XR and later, iPad Air 3+, iPad mini 5+). Chrome: Any device running Chrome browser (extension only).
Absolutely. MinuteAI never uploads your audio or transcripts to our servers. All processing happens on your device. Optional iCloud sync is end-to-end encrypted and managed by Apple, not us.
Yes. All transcripts are editable within the app. You can also export to Markdown or TXT and edit in your favorite text editor.
Yes. Cancel anytime from your Apple ID settings. No cancellation fees. Your Pro features remain active until the end of your billing period.
Start Transcribing in 60 Seconds
Free forever. No credit card. No setup hassle. Just download and start recording.