# How loqy is built and why.

> Engineering notes, product principles, local AI model evaluations, and major loqy milestones.

Inside loqy

Models, architecture, and the decisions shaping loqy.

[Atom feed](https://loqy.work/atom.xml)

Featured

Article 01 Principles August 12, 2026

## [Why loqy starts local and how cloud can complement it](https://loqy.work/blog/why-loqy-starts-local/)

The engine and work stay local; choose the model for each task.

[Read article](https://loqy.work/blog/why-loqy-starts-local/)

## The series

Article 02 Architecture August 13, 2026

### [Where AI agent work actually runs](https://loqy.work/blog/where-ai-agent-work-runs/)

One request crosses distinct zones; each receives only the data and permissions it needs.

[Read article](https://loqy.work/blog/where-ai-agent-work-runs/)

Article 03 Economics August 13, 2026

### [What local and hosted AI really costs](https://loqy.work/blog/cost-of-local-and-hosted-ai/)

Compare local and hosted inference, from the device to the accepted deliverable.

[Read article](https://loqy.work/blog/cost-of-local-and-hosted-ai/)

Article 04 Ownership August 13, 2026

### [Own your work: portability, open formats, and freedom to leave](https://loqy.work/blog/own-your-work/)

Work stays under your control when you can read it, resume it, and take it elsewhere.

[Read article](https://loqy.work/blog/own-your-work/)

Article 05 Sovereignty August 13, 2026

### [What it means to build a European AI agent](https://loqy.work/blog/building-a-european-ai-agent/)

Connecting origin, processing locations, formats, and exit paths makes sovereignty assessable.

[Read article](https://loqy.work/blog/building-a-european-ai-agent/)

Article 06 Model evaluation August 12, 2026

### [How loqy evaluates local AI models for real work](https://loqy.work/blog/how-loqy-evaluates-local-models/)

A local model is kept only if it can finish real work on the target Mac.

[Read article](https://loqy.work/blog/how-loqy-evaluates-local-models/)

Article 07 Evidence August 12, 2026

### [Why AI agents need evidence](https://loqy.work/blog/why-ai-agents-need-evidence/)

Connecting sources to decisions makes every result verifiable.

[Read article](https://loqy.work/blog/why-ai-agents-need-evidence/)