Amd Lemonade Local Ai Adds Mcp Gateway For App Integration
The MCP gateway turns Lemonade from a standalone local AI runner into a backend that cutting-edge cloud AI agents can offload privacy-sensitive or computationally heavy tasks to, keeping them local and free.
Reporting from 1 sources: GIGAZINE.
AMD released version 10.8.0 of its local AI execution environment Lemonade on June 18, 2026, adding an MCP gateway feature. This allows MCP-compatible apps like Claude Code and GitHub Copilot to call AI models running on Lemonade via an MCP server, enabling tasks such as batch classification, transcription, and image generation to run locally and offline.
AMD released version 10.8.0 of its local AI execution environment Lemonade on June 18, adding an MCP gateway feature. The update lets MCP-compatible apps like Claude Code, GitHub Copilot, Claude Desktop, Cursor, and Cline call AI models running on Lemonade through an MCP server. According to the pull request, agents on cutting-edge cloud models can route privacy-sensitive or computationally heavy tasks-such as batch classification, transcription, and image generation-to local AI models without interrupting the conversation, executing them for free and offline on the user's chosen model. Lemonade runs on Windows, macOS, and Linux, and is optimized for AMD GPUs, CPUs, and NPUs while also supporting NVIDIA GPUs.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.