Installation
System Requirements
| Component | Minimum | Recommended |
|---|---|---|
| CPU | 4 cores (AMD64 or ARM64) | 8+ cores (AVX-512 preferred) |
| RAM | 8 GB | 16 GB+ |
| Storage | 5 GB SSD | 50 GB NVMe |
| OS | Linux (Kernel 5.4+), macOS 11+, Windows (WSL2) | Linux (Ubuntu 22.04+) |
Install
- Linux
- macOS
- Windows (WSL2)
Python
pip install cortex-runtime==2.5.10
Node.js
npm install @dopove/cortex@latest
CLI only
curl -fsSL https://get.dopove.com/cortex | sh
Python (Apple Silicon and Intel)
pip install cortex-runtime==2.5.10
Node.js
npm install @dopove/cortex@latest
CLI only
curl -fsSL https://get.dopove.com/cortex | sh
Run inside your WSL2 terminal:
pip install cortex-runtime==2.5.10
# or
npm install @dopove/cortex@latest
Verify
cortex doctor
Output should confirm runtime health, architecture detection, and dependency resolution:
✔ Runtime: cortex-runtime v2.5.10 (STABLE)
✔ Engine: Rust-native (Tokio v1.35)
✔ Architecture: arm64 (Turbo Mode Enabled)
✔ Shared KV Cache: Active
✔ Python 3.12: found (/usr/bin/python3)
✔ cgroups v2: available (VPC Isolation Active)
✔ Namespace isolation: available
Tier Comparison
| Feature | Community | Developer Pro | Enterprise |
|---|---|---|---|
| Cost | Free | $9/mo | Custom |
| Max Concurrent Agents | 10 | 100 | Node-scaled |
| Architecture | Generic | Native (ARM64/AMD64) | Native + AVX-512 |
| Turbo Mode | Basic | 3x Parallel | 5x Parallel |
| Support | Community | Priority | Dedicated |
Enterprise
For high-concurrency, air-gapped, or K8s deployments, contact us.