Phone as interface, home servers as the lab: a two-workflow local AI setup
EAccelerate_42 · x · 2026-10-16
The author shows two ways he accesses his own local GPU servers (DGX/GPU boxes) from his phone via ModelDeck:
- Direct chat: phone connects privately straight to local model servers — models run entirely on his own hardware.
- Coding agent: the phone fronts a coding agent running on a Mac, which in turn taps the local model servers.
Core idea: "the phone is the interface, the lab runs the models" — decoupling the mobile client from self-hosted compute for privacy and horsepower.
More from coding & agent
- Dev uses Grok to clean up 3D printer firmware, catches it sneaking in nonsense code — yacineMTB · 2026-10-16
- Guardrails can make risky delegation work — but it takes a full-time team, says iandanforth — iandanforth · 2026-10-16
- What can you truly delegate? Only tasks with unambiguous, verifiable results — iandanforth · 2026-10-16
- Yacine: vibe-coded repos with 3,330 agent-written tests are all useless — yacineMTB · 2026-10-16
- Dev argues coding models now beat nearly all engineers in every branch of software — zack_overflow · 2026-10-16
- Q4-quantized coder model on 8-core CPU builds interactive solar system demo in one shot — IntrepidMindExplorer · 2026-10-16