---
title: "copilot-local-llm"
description: "Discover and register models from supported local LLM providers in Copilot sessions."
canonical: https://agentpluginsdirectory.com/plugins/copilot-local-llm
last-updated: 2026-10-05
---

# copilot-local-llm
Discover and register models from supported local LLM providers in Copilot sessions.
- Slug: copilot-local-llm
- Publisher: Mike Wallio
- Repository: https://github.com/scaryrawr/scarypilot
- Manifest: plugins/copilot-local-llm/plugin.json
- Version: 0.2.6
- Category (editorial): other
- Skills: 0
- MCP servers: 0
- Stars: 5
- Repository created: 2026-02-12
- Repository last pushed: 2026-10-03
- Publisher type: User
- Listing: https://agentpluginsdirectory.com/plugins/copilot-local-llm
- Schema: https://agent-plugins.org/schemas/1.0.0/plugin.schema.json

## What copilot-local-llm does, in the publisher's words

copilot-local-llm exposes models from local OpenAI-compatible servers in the Copilot model picker. It adds providers only when their model-discovery endpoint is available, so Copilot-hosted models continue to work normally. Discovery requests time out after three seconds to bound extension startup time. Discovered models are registered as part of the session join, so resuming the CLI does not replay additive provider registrations. Models remain available through /model as provider/model-id (for example, omlx/Qwen3.5-9B-mxfp4).

The plugin does not modify Copilot's system prompt or available tools. For a local-model launcher with minimal prompts and tooling, see vivi.

- A current GitHub Copilot CLI release with plugin extension support.
- At least one supported local model server. Unavailable servers are ignored.

From the project README, punctuation lightly normalized. Full text: https://raw.githubusercontent.com/scaryrawr/scarypilot/HEAD/plugins/copilot-local-llm/README.md