Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.bricks.tools/llms.txt

Use this file to discover all available pages before exploring further.

BRICKS Buttress

Preview

Heavy models, on the right hardware.

BRICKS Buttress is a backend service that offloads compute-intensive generator work — LLM inference, speech-to-text — from BRICKS Foundation devices to a more powerful machine on the same LAN. When a tablet or kiosk can’t run a large model locally, it transparently delegates the work to a Buttress server while keeping the same generator API.

Features

Quick start

When to use Buttress

Good fit

  • Models too large to fit on the device
  • Many devices on the same LAN sharing one model
  • Devices that need fast first-token latency
  • Battery-sensitive mobile deployments

Not needed

  • Small models that already run well on-device
  • Air-gapped devices with no LAN connectivity
  • Single-device prototypes or demos