LWLLM Worker
Sign in

Cloud & local proxy for your AI coding

Route models.
Keep the keys.

LLM Worker is the control plane for your model gateway. Manage provider credentials, catalogs, budgets, and remote machines — then let a cloud or local proxy route every request from VS Code with retries and fallbacks.

Get Started See how LLM Worker works

Closed beta preview — GitHub authentication required.

Capabilities

One surface for routing, keys, and machines

Cloud proxy, managed

A managed gateway with admission control, automatic retries, and fallback routing across your providers.

Local proxy, no binaries

An embedded JavaScript runtime runs a local proxy inside VS Code — no native executables to trust.

Remote Cockpit

Connect remote machines to the control plane and operate them from one workspace surface.

Bring your own keys

Provider credentials stay under your control — OpenAI-compatible providers plug in without vendor lock-in.

Telemetry & budgets

Usage, routing, and spend signals per provider, with budget guardrails on top.

Private by design

Provider-agnostic routing. Prompt content never enters telemetry.

How it works

From VSIX to first message in three steps

  1. 01

    Install the VSIX

    Drop the LLM Worker extension into VS Code — the local proxy runtime is embedded, nothing else to install.

  2. 02

    Connect GitHub

    Authenticate with GitHub to link your editor to your workspace. Your keys stay with you.

  3. 03

    Chat

    Start a conversation. Every request is routed through your cloud or local gateway with retries and fallbacks applied.

Get Started
LWLLM Worker

Closed beta — join now

LLM Worker is in closed beta. Your workspace is isolated and provisioned on sign-in — expect breaking changes between builds while we iterate.

Join the beta