{
  "id": 15754741,
  "title": "I remapped the Copilot key on my keyboard to open my local model instead",
  "url": "https://urgent.news/2026/10/20/i-remapped-the-copilot-key-on-my-keyboard-to-open-my-local-model",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-20T14:30:19.000Z",
  "source": {
    "name": "XDA Developers",
    "slug": "xda-developers",
    "url": "https://www.xda-developers.com/remapped-copilot-key-keyboard-open-local-model-instead/"
  },
  "original_language": "en",
  "account": "The Microsoft Copilot feature, initially intended to simplify AI accessibility for users with limited technical expertise, has become a source of frustration for many users, including the author of this article. The author's Lenovo Yoga Slim 7x Gen 9 laptop, a Copilot+ PC equipped with a Snapdragon X Elite CPU, is capable of running local models. By utilizing PowerToys and performing some modifications, the author was able to remap the dedicated Copilot button on their keyboard to launch a local LLM instead. This change allowed the author to leverage the powerful capabilities of the Snapdragon X Elite processor, specifically its integrated NPU, which operates at 45 TOPS.\n\nThe Snapdragon X Elite chip, an Arm-based processor, enabled the author to test the performance of a locally hosted LLM. Despite having only 16GB of RAM, a small yet effective model, such as qwen3.5:4b, could run smoothly on the device. The author chose Qwen3.5:4b for testing, which fits within the memory constraints of the laptop. The author emphasized the importance of running LLMs on local hardware to maintain control over prompts, replies, and processing. This approach aligns with their preference for storing files on a NAS and using open-weight models without relying on remote inference.\n\nThe author expressed satisfaction with the results after remapping the key, as the AI key now launched LM Studio, loading the desired local Qwen model. Although there is a slight delay when the key is pressed due to RAM allocation for the model, the author grew accustomed to this delay when using self-hosted AI tools. However, the author noted that enabling or disabling PowerToys is necessary for the remap to function, and the key can still be used to launch Microsoft's Copilot. The author questioned why the dedicated AI key is tied to Microsoft's Copilot, a cloud-based AI service, rather than allowing users to choose any local LLM.\n\nThe author concluded that while Microsoft is developing its own LLM based on OpenAI's GPT, the dedicated Copilot key on their keyboard still opens a cloud-based service. The author believes that users should have the freedom to decide how they want to use AI on their own Windows PC. The author's experience with the Lenovo laptop highlights the limitations of running local models, such as the constraints imposed by the 16GB RAM and the lack of a GPU. Despite these limitations, the author is confident that running LLMs locally offers benefits, including greater control, cost savings, and avoiding subscriptions to large tech companies.",
  "summary": "Begone, Microslop!",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}