Build. Deploy. Manage.
Enterprise Agents

The open platform that accelerates agent adoption. From discovering what to automate,
to launching your first agent, to optimizing your thousandth.

100% Local & Private
MIT Open Source License
0 Telemetry / Tracking
Agent Workflows

Know What to Automate

CortexOS diagnostics give you complete visibility into your hardware capabilities and AI potential before you build a single workflow.

Hardware Diagnostics

Real-time CPU, RAM, and VRAM profiling. CortexOS knows your stack and automatically recommends optimal model sizes and concurrency settings.

AutoModel Sizing

Agent Cookbook

Pre-built agent recipes for common enterprise workflows. Research, document generation, email triage, calendar management — ready in seconds.

50+Agent Templates

Deep Research Mode

Multi-step autonomous research agents that crawl, summarize, and synthesize information across documents, web, and internal knowledge bases.

Multi-hopReasoning Chains

Gateway Triggers

Webhook, cron, email, and calendar triggers route work to the right agents automatically. Natural language scheduling — "every Monday at 9am".

NLCron Scheduling

Multi-Agent Workflows, Your Way

Chain intelligent agents together with CortexOS's visual pipeline builder. Connect Chat, Documents, Memory, and Skills into powerful automated systems.

🤖

Autonomous Agents

Full autonomous agent panel with tool-use, memory, and multi-step planning.

💬

Chat with Branching

Branch conversations, compare responses, and maintain full conversation history.

📄

Documents + AI Copilot

Rich text editor with inline AI assistance, summarization, and drafting.

🧠

Memory & Skills

Persistent vector memory and reusable skill modules across all your agents.

Scheduled Automations

Natural language cron, email/CalDAV integration, and webhook pipelines.

🔗

API & Integrations

Connect to internal tools, databases, and external services via OpenAPI specs.

Enterprise-Grade Control

Govern every agent, workflow, and user with RBAC, audit logs, sandbox isolation, and compliance-ready tooling — all running on your infrastructure.

🔒

Sandbox Isolation

Choose your execution environment: Docker, SSH, or native. Seccomp hardening and namespace isolation by default.

DockerSSHSeccomp
👤

RBAC & Audit Logs

Fine-grained role-based access control with immutable audit trails for every agent action and user decision.

RBACSOC2-ReadyGDPR
📅

Calendar & Email Integration

CalDAV and SMTP/IMAP integration. Agents read your calendar, send emails, and manage tasks on your behalf.

CalDAVSMTPIMAP
⏱️

Natural Language Cron

Schedule complex multi-agent workflows in plain English. "Run the daily report every weekday at 8am" — just works.

NL SchedulingCron

Continuous Intelligence Improvement

Full observability stack built in. Monitor, compare, and improve every model and agent using private data — no data leaves your servers.

🧬

Memory Vector Timeline

Visualize your agent's memory graph over time. Inspect, prune, and curate what your agents know and remember.

Week 1
Week 2
Week 3
Week 4
This week
📊

RL Training Metrics

Fine-tune models with Reinforcement Learning from Human Feedback entirely on-prem. Track reward curves and loss metrics in real time.

Reward+0.87
Loss0.023
Steps12.4k
Accuracy94.2%
⚖️

Blind Comparison Testing

A/B test any two models or prompts side-by-side with blind human scoring. Find what actually works best for your specific use case.

Model A
72%
Model B ✓
91%
🔭

Full Observability

Trace every agent call, tool use, and token. Export metrics to Prometheus, Grafana, or any OpenTelemetry-compatible collector.

OpenTelemetry Prometheus Grafana Jaeger Token Tracing

Proven Local Power

Everything runs on your hardware. No cloud dependency, no vendor lock-in.

🖥️
100%
Runs on your hardware
From a Raspberry Pi to a GPU cluster
🦙
Ollama
+ vLLM + llama.cpp
Native local inference backends
📱
PWA
Installable on any device
Mobile-first, offline-ready
🔌
Hybrid
Offline + full-stack modes
OpenRouter fallback for when you need it
🛡️
Zero
Telemetry & tracking
Your data never leaves your network
⚙️
1-Click
Docker Compose deploy
Or bare-metal with start.sh
Open Source Core

Powered by CortexOS

Fully open source under MIT License. No strings attached. Self-host everything, fork it, modify it — it's yours.

⚛️
React + TypeScript
🐍
FastAPI + Python
🗄️
SQLite + Vector DB
🦙
Ollama / vLLM
🐳
Docker Compose
📦
MIT License
Privacy-first — no telemetry, no tracking, ever
Local-first inference: Ollama, vLLM, llama.cpp, OpenRouter fallback
Chat with branching & conversation history
Autonomous multi-agent workflows
Documents editor with inline AI copilot
Memory/Skills vector store per user and agent
Email & CalDAV integration (SMTP, IMAP, CalDAV)
Tasks & scheduled automations with NL cron
RL Training on private data, on-prem
Themes: Cyberpunk, Jank Mode & more
Fully responsive PWA — install on any device
One-click Docker Compose or start.sh deployment
curl -fsSL https://cortexcloud.online/install.sh | bash
git clone https://github.com/jonahthan433/CortexOS
cd CortexOS && docker compose up -d
git clone https://github.com/jonahthan433/CortexOS
cd CortexOS && ./start.sh

Ready for Enterprise?

Get managed deployment, dedicated support, SLA guarantees, compliance assistance, custom integrations, and SSO — all while keeping your data fully on-premises.

Community
Free/forever
  • Full CortexOS open source
  • Self-hosted, unlimited agents
  • Community support (GitHub)
  • All core modules included
  • MIT License
Get Started Free
Air-Gapped
Contactus
  • Everything in Enterprise
  • Fully air-gapped deployment
  • On-site engineering support
  • Government/defense ready
  • Custom security audit
Contact Sales

Deploy CortexOS Today

Start with the open-source core for free, or let us run the enterprise deployment for you.
Your AI. Your hardware. Your control.