[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-70529":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":9,"totalLinesOfCode":9,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":9,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":9,"rankLanguage":9,"license":9,"archived":22,"fork":22,"defaultBranch":23,"hasWiki":22,"hasPages":22,"topics":24,"createdAt":9,"pushedAt":9,"updatedAt":44,"readmeContent":45,"aiSummary":46,"trendingCount":16,"starSnapshotCount":16,"syncStatus":47,"lastSyncTime":48,"discoverSource":49},70529,"lemonade","lemonade-sdk\u002Flemonade","lemonade-sdk","Lemonade helps users discover and run local AI apps by serving optimized LLMs right from their own GPUs and NPUs. Join our discord: https:\u002F\u002Fdiscord.gg\u002F5xXzkMu8Zk",null,"https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade","C++",4287,336,27,263,0,68,107,383,204,29.58,false,"main",[25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43],"amd","llama","llm","llm-inference","local-server","mistral","npu","onnxruntime","qwen","openai-api","mcp","mcp-server","gpu","radeon","ryzen","vulkan","ai","genai","rocm","2026-06-12 02:02:34","## 🍋 Lemonade: Refreshingly fast local AI\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002F5xXzkMu8Zk\">\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDiscord-7289DA?logo=discord&logoColor=white\" alt=\"Discord\" \u002F>\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Fblob\u002Fmain\u002Fdocs\u002Fdev\u002Fcontribute.md\" title=\"Contribution Guide\">\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPRs-welcome-brightgreen.svg\" alt=\"PRs Welcome\" \u002F>\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Freleases\u002Flatest\" title=\"Download the latest release\">\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fv\u002Frelease\u002Flemonade-sdk\u002Flemonade?include_prereleases\" alt=\"Latest Release\" \u002F>\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Ftooomm.github.io\u002Fgithub-release-stats\u002F?username=lemonade-sdk&repository=lemonade\">\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fdownloads\u002Flemonade-sdk\u002Flemonade\u002Ftotal.svg\" alt=\"GitHub downloads\" \u002F>\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Fissues\">\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fissues\u002Flemonade-sdk\u002Flemonade\" alt=\"GitHub issues\" \u002F>\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Fblob\u002Fmain\u002FLICENSE\">\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-Apache-yellow.svg\" alt=\"License: Apache\" \u002F>\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fstar-history.com\u002F#lemonade-sdk\u002Flemonade\">\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FStar%20History-View-brightgreen\" alt=\"Star History Chart\" \u002F>\u003C\u002Fa>\n\u003C\u002Fp>\n\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Fassets\u002Fblob\u002Fmain\u002Fdocs\u002Fbanner_02.png?raw=true\" alt=\"Lemonade Banner\" \u002F>\n\u003C\u002Fp>\n\u003Ch3 align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html\">Download\u003C\u002Fa> |\n  \u003Ca href=\"https:\u002F\u002Flemonade-server.ai\u002Fdocs\u002F\">Documentation\u003C\u002Fa> |\n  \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002F5xXzkMu8Zk\">Discord\u003C\u002Fa>\n\u003C\u002Fh3>\n\nLemonade is the local AI server that gives you the same capabilities as cloud APIs, except 100% free and private. Use the latest models for chat, coding, speech, and image generation on your own NPU and GPU.\n\nLemonade comes in two flavors:\n\n* **Lemonade Server** installs a service you can connect to hundreds of great apps using standard OpenAI, Anthropic, and Ollama APIs.\n* **Embeddable Lemonade** is a portable binary you can package into your own application to give it multi-modal local AI that auto-optimizes for your user’s PC.\n\n*This project is built by the community for every PC, with optimizations by AMD engineers to get the most from Ryzen AI, Radeon, and Strix Halo PCs.*\n\n## Getting Started\n\n1. **Install**: [Windows](https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html#windows) · [Linux](https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html#linux) · [macOS](https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html#macos) · [Docker](https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html#docker) · [Source](.\u002Fdocs\u002Fdev-getting-started.md)\n2. **Get Models**: Browse and download with the [Model Manager](#model-library)\n3. **Generate**: Try models with the built-in interfaces for chat, image gen, speech gen, and more\n4. **Mobile**: Take your lemonade to go: [iOS](https:\u002F\u002Fapps.apple.com\u002Fus\u002Fapp\u002Flemonade-mobile\u002Fid6757372210) · [Android](https:\u002F\u002Fplay.google.com\u002Fstore\u002Fapps\u002Fdetails?id=com.lemonade.mobile.chat.ai&pli=1) · [Source](https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade-mobile)\n5. **Connect**: Use Lemonade with your [favorite apps](https:\u002F\u002Flemonade-server.ai\u002Fmarketplace):\n\n\u003C!-- MARKETPLACE_START -->\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Flemonade-server.ai\u002Fdocs\u002Fserver\u002Fapps\u002Fclaude-code\u002F\" title=\"Claude Code\">\u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Flemonade-sdk\u002Fmarketplace\u002Fmain\u002Fapps\u002Fclaude-code\u002Flogo.png\" alt=\"Claude Code\" width=\"60\" \u002F>\u003C\u002Fa>  \u003Ca href=\"https:\u002F\u002Fquickthoughts.ca\u002Fposts\u002Ffirefox-chatback-lemonade-sdk\u002F\" title=\"Firefox Chatbot\">\u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Flemonade-sdk\u002Fmarketplace\u002Fmain\u002Fapps\u002Ffx-chatbot\u002Flogo.png\" alt=\"Firefox Chatbot\" width=\"60\" \u002F>\u003C\u002Fa>  \u003Ca href=\"https:\u002F\u002Flemonade-server.ai\u002Fdocs\u002Fserver\u002Fapps\u002FanythingLLM\u002F\" title=\"AnythingLLM\">\u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Flemonade-sdk\u002Fmarketplace\u002Fmain\u002Fapps\u002Fanythingllm\u002Flogo.png\" alt=\"AnythingLLM\" width=\"60\" \u002F>\u003C\u002Fa>  \u003Ca href=\"https:\u002F\u002Fmarketplace.dify.ai\u002Fplugins\u002Flanggenius\u002Flemonade\" title=\"Dify\">\u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Flemonade-sdk\u002Fmarketplace\u002Fmain\u002Fapps\u002Fdify\u002Flogo.png\" alt=\"Dify\" width=\"60\" \u002F>\u003C\u002Fa>  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Famd\u002Fgaia?tab=readme-ov-file#getting-started-guide\" title=\"GAIA\">\u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Flemonade-sdk\u002Fmarketplace\u002Fmain\u002Fapps\u002Fgaia\u002Flogo.png\" alt=\"GAIA\" width=\"60\" \u002F>\u003C\u002Fa>  \u003Ca href=\"https:\u002F\u002Fadmcpr.com\u002Flocal-github-copilot-with-lemonade-server-on-windows\" title=\"GitHub Copilot\">\u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Flemonade-sdk\u002Fmarketplace\u002Fmain\u002Fapps\u002Fgithub-copilot\u002Flogo.png\" alt=\"GitHub Copilot\" width=\"60\" \u002F>\u003C\u002Fa>  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Finfinity-arcade\" title=\"Infinity Arcade\">\u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Flemonade-sdk\u002Fmarketplace\u002Fmain\u002Fapps\u002Finfinity-arcade\u002Flogo.png\" alt=\"Infinity Arcade\" width=\"60\" \u002F>\u003C\u002Fa>  \u003Ca href=\"https:\u002F\u002Fn8n.io\u002Fintegrations\u002Flemonade-model\u002F\" title=\"n8n\">\u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Flemonade-sdk\u002Fmarketplace\u002Fmain\u002Fapps\u002Fn8n\u002Flogo.png\" alt=\"n8n\" width=\"60\" \u002F>\u003C\u002Fa>  \u003Ca href=\"https:\u002F\u002Flemonade-server.ai\u002Fdocs\u002Fserver\u002Fapps\u002Fopen-webui\u002F\" title=\"Open WebUI\">\u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Flemonade-sdk\u002Fmarketplace\u002Fmain\u002Fapps\u002Fopen-webui\u002Flogo.png\" alt=\"Open WebUI\" width=\"60\" \u002F>\u003C\u002Fa>  \u003Ca href=\"https:\u002F\u002Flemonade-server.ai\u002Fdocs\u002Fserver\u002Fapps\u002Fopen-hands\u002F\" title=\"OpenHands\">\u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Flemonade-sdk\u002Fmarketplace\u002Fmain\u002Fapps\u002Fopenhands\u002Flogo.png\" alt=\"OpenHands\" width=\"60\" \u002F>\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\u003Cem>Want your app featured here? \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Fmarketplace\">Just submit a marketplace PR!\u003C\u002Fa>\u003C\u002Fem>\u003C\u002Fp>\n\u003C!-- MARKETPLACE_END -->\n\n## Supported Platforms\n\n| Platform | Build |\n|----------|-------|\n| [![Arch Linux](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FArch%20Linux-supported-1793D1?logo=arch-linux&logoColor=white)](https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html#arch) | [![Build on Arch](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002Flemonade-sdk\u002Flemonade\u002Flinux_distro_builds.yml?branch=main&label=Build%20on%20Arch)](https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Factions\u002Fworkflows\u002Flinux_distro_builds.yml) |\n| [![Debian Trixie+](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDebian-Trixie%2B-A81D33?logo=debian&logoColor=white)](https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html#debian) | [![Build on Debian](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002Flemonade-sdk\u002Flemonade\u002Flinux_distro_builds.yml?branch=main&label=Build%20on%20Debian)](https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Factions\u002Fworkflows\u002Flinux_distro_builds.yml) |\n| [![Docker](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDocker-supported-2496ED?logo=docker&logoColor=white)](https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html#docker) | [![Build Container Image](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002Flemonade-sdk\u002Flemonade\u002Fbuild-and-push-container.yml?branch=main&label=Build%20Container%20Image)](https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Factions\u002Fworkflows\u002Fbuild-and-push-container.yml) |\n| [![Fedora 43+](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FFedora-43%2B-294172?logo=fedora&logoColor=white)](https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html#fedora) | [![Build .rpm](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002Flemonade-sdk\u002Flemonade\u002Fcpp_server_build_test_release.yml?branch=main&label=Build%20.rpm)](https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Factions\u002Fworkflows\u002Fcpp_server_build_test_release.yml) |\n| [![macOS](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FmacOS-supported-999999?logo=apple&logoColor=white)](https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html#macos) | [![Build .pkg](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002Flemonade-sdk\u002Flemonade\u002Fcpp_server_build_test_release.yml?branch=main&label=Build%20.pkg)](https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Factions\u002Fworkflows\u002Fcpp_server_build_test_release.yml) |\n| [![Snap](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FSnap-supported-82BEA0?logo=snapcraft&logoColor=white)](https:\u002F\u002Fsnapcraft.io\u002Flemonade-server) | [![Build Snap](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002Flemonade-sdk\u002Flemonade-server-snap\u002Fsnap-build.yaml?branch=main&label=Build%20Snap)](https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade-server-snap\u002Factions\u002Fworkflows\u002Fsnap-build.yaml) |\n| [![Ubuntu 24.04+](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FUbuntu-24.04%2B-E95420?logo=ubuntu&logoColor=white)](https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html#ubuntu) | [![Build Launchpad PPA](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002Flemonade-sdk\u002Flemonade\u002Flaunchpad-ppa.yml?branch=main&label=Build%20Launchpad%20PPA)](https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Factions\u002Fworkflows\u002Flaunchpad-ppa.yml) |\n| [![Windows 11](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FWindows-11-0078D6?logo=windows&logoColor=white)](https:\u002F\u002Flemonade-server.ai\u002Finstall_options.html#windows) | [![Build .msi](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002Flemonade-sdk\u002Flemonade\u002Fcpp_server_build_test_release.yml?branch=main&label=Build%20.msi)](https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Factions\u002Fworkflows\u002Fcpp_server_build_test_release.yml) |\n\n## Using the CLI\n\nTo run and chat with Gemma:\n\n```\nlemonade run Gemma-4-E2B-it-GGUF\n```\n\nTo code with Lemonade models:\n\n```\nlemonade launch claude\n```\n\nMulti-modality:\n\n```\n# image gen\nlemonade run SDXL-Turbo\n\n# speech gen\nlemonade run kokoro-v1\n\n# transcription\nlemonade run Whisper-Large-v3-Turbo\n```\n\nTo see available models and download them:\n\n```\nlemonade list\n\nlemonade pull Gemma-4-E2B-it-GGUF\n```\n\nTo see the backends available on your PC:\n\n```\nlemonade backends\n```\n\n\n## Model Library\n\n\u003Cimg align=\"right\" src=\"https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Fassets\u002Fblob\u002Fmain\u002Fdocs\u002Fmodel_manager_02.png?raw=true\" alt=\"Model Manager\" width=\"280\" \u002F>\n\nLemonade supports a wide variety of LLMs (**GGUF**, **FLM**, and **ONNX**), whisper, stable diffusion, etc. models across CPU, GPU, and NPU.\n\nUse `lemonade pull` or the built-in **Model Manager** to download models. You can also import custom GGUF\u002FONNX models from Hugging Face.\n\n**[Browse all built-in models →](https:\u002F\u002Flemonade-server.ai\u002Fmodels.html)**\n\n\u003Cbr clear=\"right\"\u002F>\n\n## Supported Configurations\n\nLemonade supports multiple inference engines for LLM, speech, TTS, and image generation, and each has its own backend and hardware requirements.\n\n\u003Ctable>\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth>Modality\u003C\u002Fth>\n      \u003Cth>Engine\u003C\u002Fth>\n      \u003Cth>Backend\u003C\u002Fth>\n      \u003Cth>Device\u003C\u002Fth>\n      \u003Cth>OS\u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n    \u003Ctr>\n      \u003Ctd rowspan=\"8\">\u003Cstrong>Text generation\u003C\u002Fstrong>\u003C\u002Ftd>\n      \u003Ctd rowspan=\"5\">\u003Ccode>llamacpp\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>vulkan\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>x86_64\u003C\u002Fcode> CPU, AMD iGPU, AMD dGPU\u003C\u002Ftd>\n      \u003Ctd>Windows, Linux\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Ccode>rocm\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>Supported AMD ROCm iGPU\u002FdGPU families*\u003C\u002Ftd>\n      \u003Ctd>Windows, Linux\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Ccode>cpu\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>x86_64\u003C\u002Fcode> CPU\u003C\u002Ftd>\n      \u003Ctd>Windows, Linux\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Ccode>metal\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>Apple Silicon GPU\u003C\u002Ftd>\n      \u003Ctd>macOS\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Ccode>system\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>x86_64\u003C\u002Fcode> CPU, GPU\u003C\u002Ftd>\n      \u003Ctd>Linux\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Ccode>flm\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>npu\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>XDNA2 NPU\u003C\u002Ftd>\n      \u003Ctd>Windows, Linux\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Ccode>ryzenai-llm\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>npu\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>XDNA2 NPU\u003C\u002Ftd>\n      \u003Ctd>Windows\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Ccode>vllm\u003C\u002Fcode> (experimental)\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>rocm\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>Strix Halo iGPU (gfx1151)\u003C\u002Ftd>\n      \u003Ctd>Linux\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd rowspan=\"3\">\u003Cstrong>Speech-to-text\u003C\u002Fstrong>\u003C\u002Ftd>\n      \u003Ctd rowspan=\"3\">\u003Ccode>whispercpp\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>npu\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>XDNA2 NPU\u003C\u002Ftd>\n      \u003Ctd>Windows\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Ccode>vulkan\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>x86_64\u003C\u002Fcode> CPU\u003C\u002Ftd>\n      \u003Ctd>Linux\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Ccode>cpu\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>x86_64\u003C\u002Fcode> CPU\u003C\u002Ftd>\n      \u003Ctd>Windows, Linux\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Cstrong>Text-to-speech\u003C\u002Fstrong>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>kokoro\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>cpu\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>x86_64\u003C\u002Fcode> CPU\u003C\u002Ftd>\n      \u003Ctd>Windows, Linux\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd rowspan=\"2\">\u003Cstrong>Image generation\u003C\u002Fstrong>\u003C\u002Ftd>\n      \u003Ctd rowspan=\"2\">\u003Ccode>sd-cpp\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>rocm\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>Supported AMD ROCm iGPU\u002FdGPU families*\u003C\u002Ftd>\n      \u003Ctd>Windows, Linux\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Ccode>cpu\u003C\u002Fcode>\u003C\u002Ftd>\n      \u003Ctd>\u003Ccode>x86_64\u003C\u002Fcode> CPU\u003C\u002Ftd>\n      \u003Ctd>Windows, Linux\u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\nTo check exactly which recipes\u002Fbackends are supported on your own machine, run:\n\n```\nlemonade backends\n```\n\n\u003Cdetails>\n\u003Csummary>\u003Csmall>\u003Ci>* See supported AMD ROCm platforms\u003C\u002Fi>\u003C\u002Fsmall>\u003C\u002Fsummary>\n\n\u003Cbr>\n\n\u003Ctable>\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth>Architecture\u003C\u002Fth>\n      \u003Cth>Platform Support\u003C\u002Fth>\n      \u003Cth>GPU Models\u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n    \u003Ctr>\n      \u003Ctd>\u003Cb>gfx1151\u003C\u002Fb> (STX Halo)\u003C\u002Ftd>\n      \u003Ctd>Windows, Ubuntu\u003C\u002Ftd>\n      \u003Ctd>Ryzen AI MAX+ Pro 395\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Cb>gfx120X\u003C\u002Fb> (RDNA4)\u003C\u002Ftd>\n      \u003Ctd>Windows, Ubuntu\u003C\u002Ftd>\n      \u003Ctd>Radeon AI PRO R9700, RX 9070 XT\u002FGRE\u002F9070, RX 9060 XT\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Cb>gfx110X\u003C\u002Fb> (RDNA3)\u003C\u002Ftd>\n      \u003Ctd>Windows, Ubuntu\u003C\u002Ftd>\n      \u003Ctd>Radeon PRO W7900\u002FW7800\u002FW7700\u002FV710, RX 7900 XTX\u002FXT\u002FGRE, RX 7800 XT, RX 7700 XT\u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\u003C\u002Fdetails>\n\n## Project Roadmap\n\n| Under Development          | Under Consideration         | Recently Completed      |\n|---------------------------|-----------------------------|------------------------|\n| Native multi-modal tool calling |                              | vLLM backend            |\n| More whisper.cpp backends |                              | Port app to Tauri      |\n| More SD.cpp backends      |                              | Embeddable binary release |\n| MLX support               |                              | Image generation       |\n|                           |                              | Speech-to-text         |\n|                           |                              | Text-to-speech         |\n|                           |                              | Apps marketplace       |\n\n\n## Integrate Embeddable Lemonade in You Application\n\nEmbeddable Lemonade is a binary version of Lemonade that you can bundle into your own app to give it a portable, auto-optimizing, multi-modal local AI stack. This lets users focus on your app, with zero Lemonade installers, branding, or telemetry.\n\nCheck out the [Embeddable Lemonade guide](docs\u002Fembeddable\u002FREADME.md).\n\n\n## Connect Lemonade Server to Your Application\n\nYou can use any OpenAI-compatible client library by configuring it to use `http:\u002F\u002Flocalhost:13305\u002Fv1` as the base URL. A table containing official and popular OpenAI clients on different languages is shown below.\n\nFeel free to pick and choose your preferred language.\n\n\n| Python | C++ | Java | C# | Node.js | Go | Ruby | Rust | PHP |\n|--------|-----|------|----|---------|----|-------|------|-----|\n| [openai-python](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fopenai-python) | [openai-cpp](https:\u002F\u002Fgithub.com\u002Folrea\u002Fopenai-cpp) | [openai-java](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fopenai-java) | [openai-dotnet](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fopenai-dotnet) | [openai-node](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fopenai-node) | [go-openai](https:\u002F\u002Fgithub.com\u002Fsashabaranov\u002Fgo-openai) | [ruby-openai](https:\u002F\u002Fgithub.com\u002Falexrudall\u002Fruby-openai) | [async-openai](https:\u002F\u002Fgithub.com\u002F64bit\u002Fasync-openai) | [openai-php](https:\u002F\u002Fgithub.com\u002Fopenai-php\u002Fclient) |\n\n\n### Python Client Example\n```python\nfrom openai import OpenAI\n\n# Initialize the client to use Lemonade Server\nclient = OpenAI(\n    base_url=\"http:\u002F\u002Flocalhost:13305\u002Fapi\u002Fv1\",\n    api_key=\"lemonade\"  # required but unused\n)\n\n# Create a chat completion\ncompletion = client.chat.completions.create(\n    model=\"Gemma-4-E2B-it-GGUF\",  # or any other available model\n    messages=[\n        {\"role\": \"user\", \"content\": \"What is the capital of France?\"}\n    ]\n)\n\n# Print the response\nprint(completion.choices[0].message.content)\n```\n\nFor more detailed integration instructions, see the [Integration Guide](.\u002Fdocs\u002Fserver\u002Fserver_integration.md).\n\n## FAQ\n\nTo read our frequently asked questions, see our [FAQ Guide](.\u002Fdocs\u002Ffaq.md)\n\n## Contributing\n\nLemonade is built by the local AI community! If you would like to contribute to this project, please check out our [contribution guide](.\u002Fdocs\u002Fdev\u002Fcontribute.md).\n\n## Maintainers\n\nThis is a community project maintained by @amd-pworfolk @bitgamma @danielholanda @jeremyfowers @kenvandine @Geramy @ramkrishna2910 @sawansri @siavashhub @sofiageo @superm1 @vgodsoe, and sponsored by AMD. You can reach us by filing an [issue](https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Fissues), emailing [lemonade@amd.com](mailto:lemonade@amd.com), or joining our [Discord](https:\u002F\u002Fdiscord.gg\u002F5xXzkMu8Zk).\n\n## Code Signing Policy\n\nFree code signing provided by [SignPath.io](https:\u002F\u002Fsignpath.io), certificate by [SignPath Foundation](https:\u002F\u002Fsignpath.org).\n\n- **Committers and reviewers**: [Maintainers](#maintainers) of this repo\n- **Approvers**: [Owners](https:\u002F\u002Fgithub.com\u002Forgs\u002Flemonade-sdk\u002Fpeople?query=role%3Aowner)\n\n**Privacy policy**: This program will not transfer any information to other networked systems unless specifically requested by the user or the person installing or operating it. When the user requests it, Lemonade downloads AI models from [Hugging Face Hub](https:\u002F\u002Fhuggingface.co\u002F) (see their [privacy policy](https:\u002F\u002Fhuggingface.co\u002Fprivacy)).\n\n## License and Attribution\n\nThis project is:\n- Built with C++ (server) and React (app) with ❤️ for the open source community,\n- Standing on the shoulders of great tools from:\n  - [ggml\u002Fllama.cpp](https:\u002F\u002Fgithub.com\u002Fggml-org\u002Fllama.cpp)\n  - [ggml\u002Fwhisper.cpp](https:\u002F\u002Fgithub.com\u002Fggerganov\u002Fwhisper.cpp)\n  - [ggml\u002Fstable-diffusion.cpp](https:\u002F\u002Fgithub.com\u002Fleejet\u002Fstable-diffusion.cpp)\n  - [kokoros](https:\u002F\u002Fgithub.com\u002Flucasjinreal\u002FKokoros)\n  - [OnnxRuntime GenAI](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fonnxruntime-genai)\n  - [Hugging Face Hub](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fhuggingface_hub)\n  - [OpenAI API](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fopenai-python)\n  - [IRON\u002FMLIR-AIE](https:\u002F\u002Fgithub.com\u002FXilinx\u002Fmlir-aie)\n  - and more...\n- Licensed under the [Apache 2.0 License](https:\u002F\u002Fgithub.com\u002Flemonade-sdk\u002Flemonade\u002Fblob\u002Fmain\u002FLICENSE).\n  - Portions of the project are licensed as described in [LICENSE](.\u002FLICENSE).\n\n\u003C!--This file was originally licensed under Apache 2.0. It has been modified.\nModifications Copyright (c) 2025 AMD-->\n","Lemonade 是一个本地AI服务器，使用户能够在自己的GPU和NPU上运行优化后的大型语言模型。该项目的核心功能包括通过标准的OpenAI、Anthropic和Ollama API为用户提供聊天、编码、语音及图像生成等服务，同时支持作为嵌入式组件集成到其他应用程序中以实现多模态本地AI能力。特别适合需要在保持数据隐私的同时利用最新AI技术进行开发或研究的个人开发者与小型团队使用。项目采用C++编写，并得到了AMD工程师的支持以确保在Ryzen AI、Radeon以及Strix Halo PC上的最佳性能表现。",2,"2026-06-11 03:32:39","trending"]