🩺 Vitals
- 🟢 Last active: 2026-08-04
- 📦 Latest release: v0.63.9 (2026-07-28)
- 🐞 Open issues: 189
- 🌟 Stars: 36k
What do these metrics mean?
- Last active: when code was last pushed, as of our last check. The dot is green when that was recent, grey otherwise. A long gap can mean a tool is finished and stable, not only unmaintained.
- Latest release: the most recent tagged, packaged version the maintainers published. Not every healthy project tags releases.
- Open issues: unresolved reports and requests. A high number is normal for a popular project and is not a warning on its own.
- Stars: how many people bookmarked the project on its forge. A rough popularity signal, not a measure of quality.
🏗️ Profile
- Official: tinyhumans.ai/openhuman
- Source: github.com/tinyhumansai/openhuman
- License: GPL-3.0
- Deployment: Desktop | Local LLM (Ollama)
- Data Model: On-device SQLite + Obsidian-style Markdown vault
- Jurisdiction: UAE 🇦🇪 / Undisclosed (TinyHumans AI, founder-led)
- Compliance (SaaS): N/A (Undisclosed)
- Compliance (Self-Hosted): Self-Hosted (User Managed)
- Complexity: Low (1/5) - Native installers; runs as a standard desktop app
- Maintenance: Medium (3/5) - Fast-moving beta; single-maintainer continuity risk
- Enterprise Ready: Low (1/5) - Personal-use beta; no SSO, RBAC, audit, or certifications
1. The Executive Summary
What is it? OpenHuman is a local-first AI desktop assistant that builds a continuous, persistent memory of your work (a "second brain") and stores it entirely on your own device in an Obsidian-compatible Markdown vault and a local SQLite database. It connects to your applications and can call tools, and it can run inference on-device via Ollama, so a working setup need never send your data to a vendor cloud.
The Strategic Verdict:
- 🔴 For Enterprise / Regulated Deployment: Hard Reject (today). This is a beta-stage, single-founder personal-productivity tool with no SSO, RBAC, audit trail, or certifications. Do not sanction it as a corporate system of record or feed regulated data into it.
- 🟢 For Privacy-Conscious Knowledge Workers: Strong Buy (Conditional). If you want a persistent AI assistant whose memory you fully own (portable plain-text Markdown, local SQLite, optional offline models), OpenHuman is a credible local-first alternative to cloud assistants that retain your history on their servers.
2. The "Hidden" Costs (TCO Analysis)
| Cost Component | ChatGPT (SaaS) | OpenHuman (Self-Hosted) |
|---|---|---|
| Pricing Model | Per-seat subscription (Plus / Team / Enterprise) | Free beta (BYO LLM keys or local models) |
| Memory & Data Location | Conversation history held on vendor servers | Local Markdown vault + SQLite on your device |
| Offline Capability | None (cloud-dependent) | Optional, fully on-device via Ollama |
| Data Exit / Lock-in | Proprietary, export-limited memory | Plain-text Obsidian vault, no lock-in |
3. The "Day 2" Reality Check
🚀 Deployment & Operations
- Installation: Native installers for every desktop OS: Homebrew tap /
.dmg(macOS); signed apt repo,.deb,.AppImage, AUR (Linux);.msi(Windows). There is no server to provision or operate. - Scalability: Single-user desktop scope by design. There is no multi-user team server or central console: "scaling" means per-device installs, not centralized administration. This is a personal tool, not infrastructure.
🛡️ Security & Governance (Risk Assessment)
- Jurisdiction & Supply-Chain Trust: TinyHumans AI is a founder-led entity tied to a UAE (Dubai) jurisdiction with otherwise undisclosed corporate structure. Because OpenHuman executes locally and stores data on-device, jurisdictional data-residency exposure is largely moot. There is no vendor cloud holding your vault. The residual concern is not data location but supply-chain trust: you are accepting frequent updates from a single maintainer.
- The Compliance Shift (Endpoint Burden): As a local-first desktop app, OpenHuman holds no certifications and assumes no compliance liability. Your "second brain" is only as protected as the endpoint it lives on: full-disk encryption, device management, and backup are 100% your responsibility. There are no server-side controls because there is no server.
- License Risk (GPL-3.0 Copyleft): Released under GPL-3.0. For an individual running it as a standalone desktop assistant, the risk is minimal. For any team tempted to fork it or embed its components into proprietary internal software, the strong copyleft obligation applies. Distribution would trigger source-disclosure requirements. Treat it as an application to use, not a codebase to quietly absorb.
4. Market Landscape
🏢 Proprietary Incumbents
- ChatGPT: The dominant cloud assistant, now with persistent memory, a desktop app, and app connectors, but your history and memory live on OpenAI's infrastructure, not yours.
- Rewind AI: A personal-memory assistant that records and recalls your digital activity, but proprietary and cloud-tied, with your timeline stored off-device.
🤝 Open Source Ecosystem
- Jan: A local-first desktop assistant focused on running LLMs entirely on your own hardware. It is the closest peer on privacy, but it is a model-running chat client. It does not build OpenHuman's continuous, Obsidian-style memory vault.
- AnythingLLM: A private, local desktop app for RAG and agents over your documents. It excels at chatting with a defined corpus, whereas OpenHuman's distinction is a continuously accreting personal memory rather than workspace-scoped document retrieval.