Varven

Blog

Notes on picking and running local models, building agents and workflows, and what changes when your AI runs on your own hardware.

Local vs cloud · 2026-09-10

What actually changes when your AI runs locally

The honest tradeoffs — privacy, cost, control, and what you give up to get them.

Agents & workflows · 2026-09-05

Turning a repeated task into a scheduled run

Start with one agent, test it on its own, chain it into a workflow, then let it run without you.

Local models · 2026-09-01

A field guide to GGUF quant names

Q4_K_M, Q5_K_S, IQ2_XS — what the letters and numbers in a filename actually mean.

Local models · 2026-08-25

Picking a model that fits your machine

A plain rule of thumb for matching model size to the hardware you already have.

Three threads

Local models — matching a model’s size to your hardware, and reading a GGUF filename like you know what it says. Start with picking a model that fits your machine.

Agents & workflows — building one agent at a time, chaining them, and letting the result run on a schedule. Start with turning a repeated task into a scheduled run.

Local vs cloud — the honest tradeoffs, without picking a side for you. Start with what actually changes when your AI runs locally.