Recall everything.
Share everything.
Meetings, voice notes, and thoughts — captured, transcribed, and organized into a searchable knowledge graph. Local-first by default. Cloud when you want it.
Stop paying for six tools.Oryn replaces them all.
Everything runs locally on your Mac. No subscriptions, no accounts, no cloud dependency.
Otter.ai
Notion AI
Rewind.ai
Apple Notes
Spotlight
Voice Memos
Local‑First AI
Whisper, Qwen3‑4B, and EmbeddingGemma run on your Mac by default. Want cloud speed? Bring your own API keys — one toggle to switch.
6‑Signal Search
Hybrid search combining BM25 text ranking, vector similarity, recency decay, importance weighting, source diversity, and corroboration scoring.
Zero Setup
Download the DMG, drag to Applications, open it. No account creation needed. Works fully local out of the box.
Superpowers foryour meetings.
Eight core capabilities that turn conversations into searchable, structured knowledge — without uploading a single byte.
Knowledge Graph Memory
Every meeting builds a living knowledge graph — facts, decisions, commitments, and relationships are extracted and connected automatically.
Privacy First
Your audio never leaves your machine. All AI runs locally by default — you own 100% of your data.
Instant Summaries
Action items, decisions, and key takeaways — seconds after the call ends.
6-Signal Semantic Search
BM25 text ranking, vector similarity, recency decay, importance weighting, source diversity, and corroboration scoring — combined into a single ranked answer.
Multi‑Speaker Diarization
Knows who said what. Sherpa‑ONNX identifies speakers locally — no cloud required.
BYOK Cloud Mode
Optionally plug in your own Groq, OpenAI, or Deepgram API keys for cloud-grade speed. One toggle to switch, zero lock-in.
On-Device AI Stack
Whisper.cpp for STT, Qwen3-4B via node-llama-cpp for reasoning, EmbeddingGemma for vectors — all optimised for Apple Silicon.
Local-First Database
SQLite + FTS5 full-text search keeps everything blazing fast and completely offline.
Knowledge Graph Memory
Every meeting builds a living knowledge graph — facts, decisions, commitments, and relationships are extracted and connected automatically.
Privacy First
Your audio never leaves your machine. All AI runs locally by default — you own 100% of your data.
Instant Summaries
Action items, decisions, and key takeaways — seconds after the call ends.
6-Signal Semantic Search
BM25 text ranking, vector similarity, recency decay, importance weighting, source diversity, and corroboration scoring — combined into a single ranked answer.
Multi‑Speaker Diarization
Knows who said what. Sherpa‑ONNX identifies speakers locally — no cloud required.
BYOK Cloud Mode
Optionally plug in your own Groq, OpenAI, or Deepgram API keys for cloud-grade speed. One toggle to switch, zero lock-in.
On-Device AI Stack
Whisper.cpp for STT, Qwen3-4B via node-llama-cpp for reasoning, EmbeddingGemma for vectors — all optimised for Apple Silicon.
Local-First Database
SQLite + FTS5 full-text search keeps everything blazing fast and completely offline.
Knowledge Graph Memory
Every meeting builds a living knowledge graph — facts, decisions, commitments, and relationships are extracted and connected automatically.
Privacy First
Your audio never leaves your machine. All AI runs locally by default — you own 100% of your data.
Instant Summaries
Action items, decisions, and key takeaways — seconds after the call ends.
6-Signal Semantic Search
BM25 text ranking, vector similarity, recency decay, importance weighting, source diversity, and corroboration scoring — combined into a single ranked answer.
Multi‑Speaker Diarization
Knows who said what. Sherpa‑ONNX identifies speakers locally — no cloud required.
BYOK Cloud Mode
Optionally plug in your own Groq, OpenAI, or Deepgram API keys for cloud-grade speed. One toggle to switch, zero lock-in.
On-Device AI Stack
Whisper.cpp for STT, Qwen3-4B via node-llama-cpp for reasoning, EmbeddingGemma for vectors — all optimised for Apple Silicon.
Local-First Database
SQLite + FTS5 full-text search keeps everything blazing fast and completely offline.
Ask your second brainanything.
Forget searching through folders. Just ask. Oryn finds the answer across your meetings, brain dumps, and files using hybrid semantic search.
- •Instant answers from past meetings
- •Search across meetings, brain dumps & files
- •Track how decisions evolved over time
Reclaim yourlost hours.
Stop spending time on meeting notes and searching for past decisions. Oryn handles it automatically.
Time Saved Per Week
Hours / Year
124h
Cost
$0
- No more manual meeting notes
- Instant recall of any past conversation
From chaos to clarity.
Three simple steps to reclaiming your productivity. No setup tax.
Install & Run
Download the DMG. Drag to Applications. Open it. No account creation required. We believe in zero-setup friction.
Record or Dump
Start a meeting recording or hit Cmd+Shift+D to brain dump your thoughts. It runs quietly in the background.
AI Organizes It
The memory engine extracts facts, connects dots, and builds your personal knowledge graph automatically.
Free. Local. Yours.
macOS (Apple Silicon recommended) • ~3 GB with AI models
Local Core.Cloud Optional.
Oryn is architected to run entirely offline on Apple Silicon. Your data never leaves your machine unless you explicitly choose to use cloud providers for extra speed.
Local by Default
Zero data retention. No training on your data. Works without internet.
Cloud by Choice (BYOK)
Bring Your Own Keys for Deepgram, Groq, or OpenAI if you need massive scale.
* Apple Silicon (M1/M2/M3) recommended for local mode