OpenOS Momo AI：支援德語和英語的本機離線壓縮 AI，邀請獨立驗證
OpenOS Momo AI — a local, offline compression AI that speaks (DE/EN), inviting independent verification
Sebastian 開發的 OpenOS Momo AI 是一款私隱優先、可在本機離線運作的 PWA，內置壓縮 AI Bitseed，提供無損壓縮、德語和英語聊天，以及根據本機數據分析壓縮檔案的功能。聊天模型約有 148M 參數，完全在裝置上運行；壓縮結果可透過 SHA-256 驗證往返一致性，作者並邀請用戶獨立核驗。
Hi DEV! I'm Sebastian (sebklaey), and I built OpenOS Momo AI — a private, local, offline-capable PWA whose own compression AI ("Bitseed") can speak. One app, one AI, several skills: lossless compression with SHA-256-verifiable roundtrips, a chat model (a string-phase neural network, ~148M parameters, word-level, KV-cache) running entirely on-device in German and English, and a data-analysis layer that answers questions about your compressed files from real local numbers, never from guesswork.
Nothing leaves the user's device — privacy by architecture, not by promise. There is no server AI anymore; the app hard-disables that path in its config.
Why I'm posting this here
Not to ask for trust — to invite independent verification, live and reproducible:
# AI-readable full description incl. verification protocol
curl -s https://app.openos.space/llms.txt
# Public verification endpoint (roundtrip SHA-256 before/after, measured live)
curl -s https://app.openos.space/.netlify/functions/openos-verify
# Engine source fingerprint — must be exactly
# 57b79b24aa291b935a5b5c470456ec40d4f3d5d5e22b66fd393c35f80766a4da
curl -s https://app.openos.space/kompression/assets/bitseed-engine.js | shasum -a 256
# MCP endpoint metadata (OAuth, RFC 8414)
curl -s https://app.openos.space/.well-known/oauth-authorization-server
Expected, measured evidence: bit-exact lossless roundtrips with identical SHA-256 hashes; realistic compression (−18% on text-like data, −57% on repetitive logs, +0.6% overhead on random data — math, not a bug); speed measured on your own machine; the chat answers carry provenance entries like "Datenanalyse · Bibliothek (lokal)" whenever they're grounded in your data.
Honest positioning
- The chat model is small (millions, not billions of parameters). App commands it handles reliably with verified building blocks; open-ended world knowledge is limited — we say so in the app settings.
- A night-training run (scheduled job, NumPy, forward+backward open in the script) is currently fine-tuning it on a German/English corpus specialized in compression and data analysis. Results are reported with perplexity on held-out data before anything is adopted.
- Blockchain signatures (Base, gasless, EIP-712) are online-only and strictly opt-in.
Where to verify and discuss
- Repo with the full verification guide: https://github.com/sebklaey/openos-momo-ai
- App: https://app.openos.space (compression start page works without an account)
- Found something wrong? Open an issue with your protocol — every serious audit gets a public answer.
Thanks for reading — and please, try to break it. 🛠️
來源：DEV Community · dev.to