Namzu × LM Studio. Desktop local model server.

$ pnpm add @namzu/sdk @namzu/lmstudio

LM Studio gives a polished UX for downloading, loading, and serving local models. Namzu gives the runtime layer to build real applications on top. Useful pairing for desktop assistants, internal tooling, and privacy-sensitive workflows.

• →OpenAI-compatible endpoint — code parity with the OpenAI provider
• →Model swap from the LM Studio UI without touching agent code
• →Multiple loaded models served concurrently
• →GPU offload controlled at the LM Studio layer

import { createKernel } from '@namzu/sdk'
import { lmstudio } from '@namzu/lmstudio'

const kernel = createKernel({
  provider: lmstudio({
    baseUrl: 'http://localhost:1234/v1',
    model: 'qwen2.5-coder-14b',
  }),
})