Private AI infrastructure. Qwen3-8B. 40k context. Metal-accelerated.
Download TARX — freePlug your hardware into the TARX mesh. Process inference jobs. Get paid.
70%
Revenue share
M1+
Apple Silicon minimum
16GB
RAM minimum
Air-gapped deployment. Custom SLA. DoD-ready.
6 MCP servers · 310 tools · Available on macOS + iOS