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.