Back to ToolHalla

Local path

Local AI setup

Choose the runtime, model format, memory plan, and privacy tradeoffs before you run models on your own machine.

Decision checkpoints

01

Start with your RAM or VRAM budget, then choose model size and quantization.

02

Prefer a runtime that fits your operating system, GPU support, and workflow.

03

Use cloud GPU when context length, concurrency, or speed makes local runs painful.