Development preview · v0.0.1Windows x64 · Source-grounded guide

Your models. Your machine.

A practical field guide to your local AI workstation. Bring an existing model, understand its memory plan, and run a conversation on your own machine.

Run your first local conversation →

Know what is ready

RegesARC 0.0.1 is a Windows x64 development preview. Its native core discovers local GGUF files, reads their metadata, plans explicit CPU/GPU placement, loads a selected model through a private Lemonade runtime, and streams saved conversations with cancellation. These are real native paths, distinct from the broader interface atlas.

The tested RegesCore 35B artifact completed generation, cancellation, recovery, unload and saved-history relaunch in the installed application. That functional qualification does not establish faster inference than LM Studio. The baseline uses ordinary layer offload. Strata execution, expert paging and large-model optimization remain research.

Load with intentionExact context, explicit GPU layers and honest memory estimates.Measure what mattersSeparate prompt processing, first-token latency, decode and quality.Understand the boundariesNative host, sandboxed renderer and an owned local runtime.Follow the agent platformRepository tools, MCP, skills and orchestration are clearly labeled plans.

Choose your path

If you have the engineering build, begin with installation and quickstart. If a model fails to fit, read memory planning before changing the artifact. If you are contributing code, start with the architecture and build instructions.

There is no published RegesARC Python package, TypeScript SDK, public server or remote-link service in this release. The future platform describes the contracts and release gates they need. Do not substitute another product’s commands for an unavailable RegesARC API.

Preview the larger workstation

The 32-view interface atlas explores discovery, agent workspaces, integrations and developer operations. It is a local design simulation, not a native feature inventory. Open the interface preview with that distinction in mind.

Built by Like a King Inc., developer Davidio.dev. These guides are grounded in the current source and release evidence, with independent upstream documentation linked where it helps.