Daedalus
The brain of your home. Inference that never phones home.
Daedalus is a standalone AI inference hub that powers every device in the Pantheon ecosystem. Running optimized open-source models on ARM silicon, it handles voice recognition, camera analytics, automation logic, and more — all within your walls. Compatible with Apple Silicon Macs for expanded compute.
- Runs 7B-13B parameter models in real time
- Apple Silicon Mac cluster support
- Automatic model optimization and quantization
- Silent fanless thermal design
Product render coming soon
Features
On-Premise AI
Run large language models, vision models, and automation engines entirely on local hardware.
ARM Silicon Optimized
Purpose-built for ARM architecture with neural engine acceleration. Optionally offload to Apple Silicon Macs.
Pantheon Hub
Central orchestrator for all Pantheon devices — Talos, Argus, and Sibyl all connect here.
Model Marketplace
Browse and install community-verified open models. Swap models without touching hardware.
Encrypted Mesh
All device-to-device communication uses end-to-end encryption over a private mesh network.
Developer API
Full REST and WebSocket APIs for custom integrations, automations, and third-party hardware.
Specifications
| Processor | ARM Cortex-A78AE, 8-core |
| NPU | Dedicated 12 TOPS neural engine |
| Memory | 16GB LPDDR5 |
| Storage | 256GB NVMe + microSD expansion |
| Connectivity | 2.5GbE, Wi-Fi 6E, Bluetooth 5.3, Thread, Zigbee |
| Power | USB-C PD, 45W |
| Dimensions | 6.3" × 6.3" × 1.8" |
Ready to take back your privacy?
Join us in building a smart home that respects you. Get notified when Pantheon products are available.