GraphCanon updated 1mo · GitHub synced 1mo · 26 views this month
Decision brief
taOS is a self-hosted AI operating system tailored for environments requiring data sovereignty and privacy. It offers unique offline-first capabilities along with multi-framework support on consumer hardware.
Good fit when
- You require an AI agent OS that prioritizes your hardware ownership and offline capabilities, ensuring your data remains private without needing continuous internet access.
- Your environment consists of diverse hardware like Orange Pi or Raspberry Pi, where you aim to deploy auto-clustering for more efficient resource utilization.
Avoid when
- You prioritize cloud-based AI services with real-time data processing over self-hosted and offline solutions.
- Your infrastructure is primarily composed of enterprise-grade servers that already support advanced clustering software, making taOS's auto-clustering less beneficial.
Observed Jul 12, 2026 · Source: enrich:decision_facts
Verify the decision
Maintenance and security
Full trust report- Maintenance
- Very active (3d since push)
- As of 1mo
- Provenance
- Not a fork · Personal account
- As of 1mo
- Security (OSV)
- No lockfile
- As of 1mo
Public GitHub metadata and optional OSV scans. Signals, not a guarantee. Trust methodology.
Install
pip install taOSSimilar tools
Same-category neighbours. No typed graph edges are catalogued for this tool yet.
Evidence and technical details
Sourced facts, taxonomy, compatibility claims, README excerpt, and machine-readable endpoints.
Overview
taOS is a self-hosted operating system designed for AI agents, featuring offline AI memory, local-first functionality, full web desktop environment, and support for various hardware including Orange Pi, Raspberry Pi, and Apple Silicon. It offers auto-clustering capabilities for consumer hardware and integrates multiple AI frameworks.
Capability facts
- CLI
- CLI entrypoint
Source: pyproject.toml:[project.scripts] · Jul 26, 2026
- Languages
- python
Source: github.language+pyproject.toml · Jul 26, 2026
Categories
Compatibility
Sourced claims from the README excerpt - not unsourced marketing copy.
Source: README excerpt (regex_v1, Jul 26, 2026)
stallers touch. Nothing is hidden behind a vendored binary; everything is plain Python, plain systemd, plain shell.Source link
Tags
README
Quick Start
Controller (server):
---
# Debian / Ubuntu / Fedora / Arch / Alpine / macOS, one-line install
curl -fsSL https://raw.githubusercontent.com/jaylfc/taOS/master/scripts/install-server.sh | sudo bash
Run without sudo to install as a user-mode systemd unit instead. The script is idempotent, safe to re-run on an existing install. Supports env-var overrides for install path, branch, and port.
Linux / macOS, one-line worker install (auto-detects headless,
service otherwise; works on a fresh Debian install or your existing box)
curl -fsSL https://raw.githubusercontent.com/jaylfc/taOS/master/scripts/install-worker.sh | sudo bash -s -- http://your-server:6969
Windows 10/11, one-line worker install (PowerShell, mirrors the
Agent Deployment
5-step wizard: pick framework → choose model → configure → deploy into an isolated container (LXC on bare metal, Docker on VPS, auto-detected). Each agent gets its own memory system (taOSmd instance), its own file storage, and its own network identity. The framework runs inside the container but taOS manages everything around it: memory, channels, secrets, model access, scheduled tasks, and inter-agent communication. This means the framework is a swappable component, not a lock-in decision.
Running taOS inside an LXC (e.g. Proxmox)? Deploying an agent creates a nested container, which an unprivileged LXC cannot do -- the kernel can't remap the nested container's filesystem, so the deploy fails with an
idmapped storage / change ownershiperror. Run the taOS LXC as privileged with nesting enabled. On Proxmox: untick Unprivileged container and set Options → Features →nesting=1(pluskeyctl=1,fuse=1), then redeploy. Bare-metal and VM installs are unaffected. (taOS detects this and surfaces the fix in the deploy error.)
The Agents app on mobile -- one tap from empty to your first deployed agent.
Supported Hardware
| Category | Hardware | Notes |
|---|---|---|
| Apple Silicon | Mac Mini, MacBook, Mac Studio, Mac Pro (M1-M5) | Ollama (Metal) or MLX backend, 8-192GB unified memory |
| ARM + Rockchip NPU | Orange Pi 5/5 Plus, Rock 5B | 6 TOPS NPU, primary SBC target |
| Raspberry Pi | Pi 4 (8GB), Pi 5 (8/16GB) | CPU-only or with accelerator HATs |
| Pi Accelerators | Hailo-10H (40T), M5Stack LLM-8850 (24T) | LLM-capable accelerators |
| NVIDIA | GTX 1050 Ti through RTX 4090/5090 | CUDA 4-32GB or Vulkan legacy |
| AMD | RX 6600 through RX 7900 XTX | ROCm 4-24GB |
| Android | Flagship phones/tablets (12-16GB) | 7-8B models at 15-30 tok/s via Termux + llama.cpp |
| iOS/iPadOS | iPad Pro M4, iPhones (6-8GB+) | Dashboard via PWA, future native worker app |
| CPU Only | Any device | Smallest quantized models |
| Mixed Cluster | All of the above combined | A Mac, a Pi, a gaming PC, and an old phone, all working together |
What the install creates on your box
Full transparency on every file, service, user, and port the installers touch. Nothing is hidden behind a vendored binary; everything is plain Python, plain systemd, plain shell.
Controller install (scripts/install-server.sh)
Run curl -fsSL https://raw.githubusercontent.com/jaylfc/taOS/master/scripts/install-server.sh | sudo bash on a fresh Debian / Ubuntu / Fedora / Arch / Alpine box to get the controller fully installed, repo cloned to ~/tinyagentos/, venv created, all deps installed, and both tinyagentos.service (port 6969) and qmd.service (port 7832) registered and started.
| Where | What |
|---|---|
/etc/systemd/system/tinyagentos.service | Main controller systemd unit. Runs uvicorn on port 6969. |
/etc/systemd/system/qmd.service | Embedding backend (embed / rerank / query ex |
For agents
This page has a .md twin and JSON over the API.