Bot-Free Dual Capture
Audio loopback captures whatever is playing through your speakers (Zoom, Google Meet, Teams) plus your microphone as two clean, isolated audio channels without bot awkwardness.
Local-First · Zero-Bot · BYO-LLM
Meeting intelligence that
Meeting intelligence that
never leaves your machine.
Jot quick fragments during your call. BeeQuill captures both sides of the conversation without inviting a bot, transcribes locally via faster-whisper, and synthesizes executive meeting notes with your chosen LLM (Ollama, Claude 3.5 Sonnet, GPT-4o, Groq).
100% Offline Capable
Local SQLite Storage
Zero Data Retention
BeeQuill — Q3 Architecture Review
Enhanced NotesTranscriptAsk
Q3 Architecture Review
Executive Decisions
Action Items
Bring any model · Swap providers in real-time
OllamaClaude 3.5 SonnetGPT-4oGroq Llama-3.3OpenRouterfaster-whisperLM StudiovLLMDeepSeek
OllamaClaude 3.5 SonnetGPT-4oGroq Llama-3.3OpenRouterfaster-whisperLM StudiovLLMDeepSeek
Industry Comparison
BeeQuill vs. The Competition
See how BeeQuill's local-first architecture compares directly to Granola, Otter.ai, Fathom, and traditional cloud tools.
| Feature / Metric | BeeQuill (beequill.com) | Granola ($1.5B Val.) | Otter.ai ($100M+ ARR) | Fathom ($30M ARR) | noteforge.ai (cloud) |
|---|---|---|---|---|---|
| Bot Intrusion | Zero Bots (WASAPI/CoreAudio) | Zero Bots | Bot Joins & Records | Bot Joins & Records | Single Mic Only |
| Dual-Channel Speaker Separation | Yes (Hardware Loopback) | Yes | Cloud Diarization | Cloud Diarization | No |
| Storage Architecture | Local SQLite (On-Device) | Cloud-Stored | Cloud-Stored | Cloud-Stored | Cloud-Stored |
| Offline AI Execution | 100% Offline (Ollama + Whisper) | No (Cloud API) | No (Cloud API) | No (Cloud API) | No |
| Swappable LLMs (BYO-LLM) | Yes (Ollama/Claude/OpenAI/Groq) | Locked to Hosted AI | Proprietary Closed AI | Proprietary Closed AI | No |
| Local MCP Server | Built-in (Claude/Cursor tools) | No | No | No | No |
| Starting Price | Free Forever ($0) / $12 Pro | $14 / month | $16.99 / month | $19 / month | $18 / month |
BeeQuill offers complete local data sovereignty, zero bot awkwardness, and the freedom to bring your own local or cloud AI models without paying inflated markups.
Engineered for Privacy
All the Meeting Intelligence.
All the Meeting Intelligence.
None of the Cloud Traps.
100% Local-First SQLite
Meetings, transcripts, and notes live in a private SQLite database on your disk. Pair with local Whisper and Ollama to run your entire meeting workflow completely offline.
Streamed AI Enhancement
Your raw meeting fragments + transcripts are merged into structured executive notes, streamed token-by-token using prompt-customizable TipTap templates.
Swappable LLM Providers
Connect Ollama, Anthropic Claude 3.5, OpenAI GPT-4o, Groq, or any OpenAI-compatible server. Swap providers mid-workflow without restarting.
Workspace Semantic Chat
Query across your entire meeting archive with LangGraph agentic reasoning. Answers cite the exact source meetings and timestamps.
Private Calendar Integration
Paste a private ICS link from Google Calendar, Outlook, or Apple Calendar — zero OAuth authorization required. Automatically pre-titles meetings right before calls start.
Early Access · BeeQuill Pro
Get BeeQuill for macOS & Windows
Join engineering leaders and executives claiming back their meeting privacy. Get product updates and early access to team workspaces.
Zero spam. 100% privacy guarantee. Unsubscribe anytime.
Under the Hood
How Bot-Free Capture Actually Works
No bots interrupting your conversation. No screen sharing prompts. Pure audio hardware capture.
01
Audio Loopback Hook
Electron intercepts system audio using WASAPI loopback on Windows and CoreAudio on macOS, recording speaker sound directly.
02
Dual Independent Channels
Your microphone and remote participants are recorded as separate audio tracks, giving clean speaker separation with zero cloud costs.
03
15-Second Segment Chunks
Audio is chunked into standalone WebM segments with complete headers, ensuring zero audio corruption even if calls terminate unexpectedly.
04
Local Whisper Transcription
Segments stream in real-time to your local faster-whisper instance, outputting transcripts directly into your local database.
Own Your Meetings.
Write with BeeQuill.
Experience local privacy, zero bots, and flexible AI models today.