Hugging Face's logo

Buckets:
hf-doc-build
/
doc-dev

Files
xet
hf-doc-build/doc-dev / hub /pr_2602 /en /agents-local.html
HuggingFaceDocBuilder's picture
HuggingFaceDocBuilder
download
raw
58.4 kB
<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Local Agents with llama.cpp&quot;,&quot;local&quot;:&quot;local-agents-with-llamacpp&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Getting Started&quot;,&quot;local&quot;:&quot;getting-started&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;1. Set Your Local Hardware&quot;,&quot;local&quot;:&quot;1-set-your-local-hardware&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;2. Find a Compatible Model&quot;,&quot;local&quot;:&quot;2-find-a-compatible-model&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;3. Launch the llama.cpp Server&quot;,&quot;local&quot;:&quot;3-launch-the-llamacpp-server&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;4. Connect Your Agent&quot;,&quot;local&quot;:&quot;4-connect-your-agent&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Pi&quot;,&quot;local&quot;:&quot;pi&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Enabling vision support&quot;,&quot;local&quot;:&quot;enabling-vision-support&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;OpenClaw&quot;,&quot;local&quot;:&quot;openclaw&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Local Memory Search for OpenClaw&quot;,&quot;local&quot;:&quot;local-memory-search-for-openclaw&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Hermes Agent&quot;,&quot;local&quot;:&quot;hermes-agent&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;OpenCode&quot;,&quot;local&quot;:&quot;opencode&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;How It Works&quot;,&quot;local&quot;:&quot;how-it-works&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Alternative: llama-agent&quot;,&quot;local&quot;:&quot;alternative-llama-agent&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Next Steps&quot;,&quot;local&quot;:&quot;next-steps&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}">
 <link href="/docs/hub/pr_2602/en/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/entry/start.cc882a17.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/chunks/scheduler.409792a1.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/chunks/singletons.cb2b08f5.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/chunks/index.0f0d9f26.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/chunks/paths.26fd1595.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/entry/app.07a28e47.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/chunks/preload-helper.ff0ae103.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/chunks/index.92d389ff.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/nodes/0.7c64a1ba.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/chunks/each.e59479a4.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/nodes/9.7094df87.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/chunks/CopyLLMTxtMenu.75899c1b.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/chunks/MermaidChart.svelte_svelte_type_style_lang.d35b6be7.js">
 <link rel="modulepreload" href="/docs/hub/pr_2602/en/_app/immutable/chunks/CodeBlock.d7fa2a05.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Local Agents with llama.cpp&quot;,&quot;local&quot;:&quot;local-agents-with-llamacpp&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Getting Started&quot;,&quot;local&quot;:&quot;getting-started&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;1. Set Your Local Hardware&quot;,&quot;local&quot;:&quot;1-set-your-local-hardware&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;2. Find a Compatible Model&quot;,&quot;local&quot;:&quot;2-find-a-compatible-model&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;3. Launch the llama.cpp Server&quot;,&quot;local&quot;:&quot;3-launch-the-llamacpp-server&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;4. Connect Your Agent&quot;,&quot;local&quot;:&quot;4-connect-your-agent&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Pi&quot;,&quot;local&quot;:&quot;pi&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Enabling vision support&quot;,&quot;local&quot;:&quot;enabling-vision-support&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;OpenClaw&quot;,&quot;local&quot;:&quot;openclaw&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Local Memory Search for OpenClaw&quot;,&quot;local&quot;:&quot;local-memory-search-for-openclaw&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Hermes Agent&quot;,&quot;local&quot;:&quot;hermes-agent&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;OpenCode&quot;,&quot;local&quot;:&quot;opencode&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;How It Works&quot;,&quot;local&quot;:&quot;how-it-works&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Alternative: llama-agent&quot;,&quot;local&quot;:&quot;alternative-llama-agent&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Next Steps&quot;,&quot;local&quot;:&quot;next-steps&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <div class="items-center shrink-0 min-w-[100px] max-sm:min-w-[50px] justify-end ml-auto flex" style="float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"><div class="inline-flex rounded-md max-sm:rounded-sm"><button class="inline-flex items-center gap-1 h-7 max-sm:h-7 px-2 max-sm:px-1.5 text-sm font-medium text-gray-800 border border-r-0 rounded-l-md max-sm:rounded-l-sm border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-live="polite"><span class="inline-flex items-center justify-center rounded-md p-0.5 max-sm:p-0 hover:text-gray-800 dark:hover:text-gray-200"><svg class="sm:size-3.5 size-3" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg></span> <span>Copy page</span></button> <button class="inline-flex items-center justify-center w-6 max-sm:w-5 h-7 max-sm:h-7 disabled:pointer-events-none text-sm text-gray-500 hover:text-gray-700 dark:hover:text-white rounded-r-md max-sm:rounded-r-sm border border-l transition border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-haspopup="menu" aria-expanded="false" aria-label="Open copy menu"><svg class="transition-transform text-gray-400 overflow-visible sm:size-3.5 size-3 rotate-0" width="1em" height="1em" viewBox="0 0 12 7" fill="none" xmlns="http://www.w3.org/2000/svg"><path d="M1 1L6 6L11 1" stroke="currentColor"></path></svg></button></div> </div> <h1 class="relative group"><a id="local-agents-with-llamacpp" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#local-agents-with-llamacpp"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Local Agents with llama.cpp</span></h1> <p data-svelte-h="svelte-qstqi4">You can run a coding agent entirely on your own hardware. Several open-source agents can connect to a local <a href="https://github.com/ggerganov/llama.cpp" rel="nofollow">llama.cpp</a> server to give you an experience similar to Claude Code or Codex — but everything runs on your machine.</p> <h2 class="relative group"><a id="getting-started" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#getting-started"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Getting Started</span></h2> <h3 class="relative group"><a id="1-set-your-local-hardware" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#1-set-your-local-hardware"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>1. Set Your Local Hardware</span></h3> <p data-svelte-h="svelte-wc2i8u">Set your local hardware so it can show you which models are compatible with your setup.</p> <p data-svelte-h="svelte-8625xd">Go to <a href="https://huggingface.co/settings/hardware" rel="nofollow">huggingface.co/settings/hardware</a> and configure your local hardware profile. Then select <code>llama.cpp</code> in your <a href="https://huggingface.co/settings/local-apps" rel="nofollow">Local Apps settings</a>, as this will be the engine you’ll use.</p> <h3 class="relative group"><a id="2-find-a-compatible-model" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#2-find-a-compatible-model"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>2. Find a Compatible Model</span></h3> <p data-svelte-h="svelte-16mkm91">Browse for <a href="https://huggingface.co/models?apps=llama.cpp&sort=trending" rel="nofollow">Llama.cpp-compatible models</a>.</p> <h3 class="relative group"><a id="3-launch-the-llamacpp-server" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#3-launch-the-llamacpp-server"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>3. Launch the llama.cpp Server</span></h3> <p data-svelte-h="svelte-x9l43b">On the model page, click the <strong>“Use this model”</strong> button and select <code>llama.cpp</code>. It will show you the exact commands for your setup. The first step is to start a llama.cpp server, e.g.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-bash "><!-- HTML_TAG_START -->llama-server -hf ggml-org/gemma-4-26b-a4b-it-GGUF:Q4_K_M<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-j5ue6h">This downloads the model and starts an OpenAI-compatible API server on your machine. See the <a href="./gguf-llamacpp">llama.cpp guide</a> for installation instructions.</p> <h3 class="relative group"><a id="4-connect-your-agent" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#4-connect-your-agent"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>4. Connect Your Agent</span></h3> <p data-svelte-h="svelte-vd90t9">Pick one of the agents below and follow the setup instructions.</p> <h2 class="relative group"><a id="pi" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#pi"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Pi</span></h2> <p data-svelte-h="svelte-13wegxu"><a href="https://pi.dev" rel="nofollow">Pi</a> is the agent behind <a href="https://github.com/openclaw" rel="nofollow">OpenClaw</a> and is now integrated directly into Hugging Face, giving you access to thousands of compatible models.</p> <p data-svelte-h="svelte-1s1rkvg">Install Pi:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-bash "><!-- HTML_TAG_START -->npm install -g @mariozechner/pi-coding-agent<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1gv7ke0">Then add your local model to Pi’s configuration file at <code>~/.pi/agent/models.json</code>:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-json "><!-- HTML_TAG_START --><span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;providers&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;llama-cpp&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;baseUrl&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-string">&quot;http://localhost:8080/v1&quot;</span><span class="hljs-punctuation">,</span>
 <span class="hljs-attr">&quot;api&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-string">&quot;openai-completions&quot;</span><span class="hljs-punctuation">,</span>
 <span class="hljs-attr">&quot;apiKey&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-string">&quot;none&quot;</span><span class="hljs-punctuation">,</span>
 <span class="hljs-attr">&quot;models&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">[</span>
 <span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;id&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-string">&quot;ggml-org-gemma-4-26b-4b-gguf&quot;</span>
 <span class="hljs-punctuation">}</span>
 <span class="hljs-punctuation">]</span>
 <span class="hljs-punctuation">}</span>
 <span class="hljs-punctuation">}</span>
<span class="hljs-punctuation">}</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1uwkzgb">Start Pi in your project directory:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-bash "><!-- HTML_TAG_START -->pi<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1vahsm2">Pi connects to your local llama.cpp server and gives you an interactive agent session.</p> <p data-svelte-h="svelte-1vlzma4"><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/hub/pi-llama-cpp-demo.gif" alt="Demo"></p> <h3 class="relative group"><a id="enabling-vision-support" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#enabling-vision-support"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Enabling vision support</span></h3> <p data-svelte-h="svelte-aq6pwa">For vision-capable models, add <code>&quot;input&quot;: [&quot;text&quot;, &quot;image&quot;]</code> to the model entry in <code>~/.pi/agent/models.json</code>:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-json "><!-- HTML_TAG_START --><span class="hljs-attr">&quot;models&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">[</span>
 <span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;id&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-string">&quot;unsloth/Qwen3.6-35B-A3B-GGUF:Q4_K_XL&quot;</span><span class="hljs-punctuation">,</span>
 <span class="hljs-attr">&quot;input&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">[</span><span class="hljs-string">&quot;text&quot;</span><span class="hljs-punctuation">,</span> <span class="hljs-string">&quot;image&quot;</span><span class="hljs-punctuation">]</span>
 <span class="hljs-punctuation">}</span>
<span class="hljs-punctuation">]</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1ubrdiv">Browse <a href="https://huggingface.co/models?pipeline_tag=image-text-to-text&apps=pi" rel="nofollow">vision-language models compatible with Pi</a>.</p> <h2 class="relative group"><a id="openclaw" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#openclaw"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>OpenClaw</span></h2> <p data-svelte-h="svelte-1qbo6rz"><a href="https://github.com/openclaw" rel="nofollow">OpenClaw</a> works locally with llama.cpp. You can set your model via the onboard command:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-bash "><!-- HTML_TAG_START -->openclaw onboard --non-interactive \
 --auth-choice custom-api-key \
 --custom-base-url <span class="hljs-string">&quot;http://127.0.0.1:8080/v1&quot;</span> \
 --custom-model-id <span class="hljs-string">&quot;ggml-org-gemma-4-26b-a4b-gguf&quot;</span> \
 --custom-api-key <span class="hljs-string">&quot;llama.cpp&quot;</span> \
 --secret-input-mode plaintext \
 --custom-compatibility openai \
 --accept-risk<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-n31xv0">You can also run <code>openclaw onboard</code> interactively, select <code>custom-compatibility</code> with <code>openai</code>, and pass the same configuration.</p> <h3 class="relative group"><a id="local-memory-search-for-openclaw" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#local-memory-search-for-openclaw"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Local Memory Search for OpenClaw</span></h3> <p data-svelte-h="svelte-7xnlxi">You can run local embedding models with Llama.cpp for your agent’s memory search. To do so, make sure to have node-llama-cpp.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-bash "><!-- HTML_TAG_START -->npm i node-llama-cpp <!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1h7l15u">Here’s an example snippet to run <a href="https://huggingface.co/ggml-org/embeddinggemma-300M-GGUF?show_file_info=embeddinggemma-300M-Q8_0.gguf" rel="nofollow">quantized EmbeddingGemma-300M</a> locally for memory search. OpenClaw automatically downloads and serves the model with the command below.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-bash "><!-- HTML_TAG_START -->openclaw config <span class="hljs-built_in">set</span> agents.defaults.memorySearch.provider <span class="hljs-built_in">local</span>
openclaw config <span class="hljs-built_in">set</span> agents.defaults.memorySearch.local.modelPath <span class="hljs-string">&quot;hf:ggml-org/embeddinggemma-300m-qat-q8_0-GGUF/embeddinggemma-300m-qat-Q8_0.gguf&quot;</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-at6ger">Restart the gateway and validate.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-bash "><!-- HTML_TAG_START -->openclaw gateway restart
openclaw memory status
<span class="hljs-comment"># Memory Search (main)</span>
<span class="hljs-comment"># Provider: local (requested: local)</span>
<span class="hljs-comment"># Model: hf:ggml-org/embeddinggemma-300m-qat-q8_0-GGUF/embeddinggemma-300m-qat-Q8_0.gguf</span><!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="hermes-agent" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#hermes-agent"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Hermes Agent</span></h2> <p data-svelte-h="svelte-12g04be"><a href="https://hermes-agent.nousresearch.com/" rel="nofollow">Hermes Agent</a> works locally with llama.cpp. Define a default config as:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-yaml "><!-- HTML_TAG_START --><span class="hljs-attr">model:</span>
 <span class="hljs-attr">provider:</span> <span class="hljs-string">custom</span>
 <span class="hljs-attr">default:</span> <span class="hljs-string">ggml-org/gemma-4-26B-A4B-it-GGUF:Q4_K_M</span>
 <span class="hljs-attr">base_url:</span> <span class="hljs-string">http://127.0.0.1:8080/v1</span>
 <span class="hljs-attr">api_key:</span> <span class="hljs-string">llama.cpp</span>
<span class="hljs-attr">custom_providers:</span>
 <span class="hljs-bullet">-</span> <span class="hljs-attr">name:</span> <span class="hljs-string">Local</span> <span class="hljs-string">(127.0.0.1:8080)</span>
 <span class="hljs-attr">base_url:</span> <span class="hljs-string">http://127.0.0.1:8080/v1</span>
 <span class="hljs-attr">api_key:</span> <span class="hljs-string">llama.cpp</span>
 <span class="hljs-attr">model:</span> <span class="hljs-string">ggml-org/gemma-4-26B-A4B-it-GGUF:Q4_K_M</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-txa203">### Local Memory Search for Hermes Agent</p> <p data-svelte-h="svelte-fuyqou">Hermes Agent consumes semantic search models through endpoints. Once you get your preferred embedding model up on endpoint 8080 with llama.cpp or the inference engine of your choice, add the following to <code>~/.hermes/config.yaml</code>.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-bash "><!-- HTML_TAG_START -->auxiliary:
 session_search:
 base_url: <span class="hljs-string">&quot;http://127.0.0.1:8080/v1&quot;</span>
 api_key: <span class="hljs-string">&quot;no-key-required&quot;</span>
 model: <span class="hljs-string">&quot;local-llama&quot;</span> <span class="hljs-comment"># your model alias</span>
 <span class="hljs-built_in">timeout</span>: 90
 max_concurrency: 1<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-13wgqa7">Check if this works, <code>none - built-in only</code> shows that no other memory plug-ins are used. The output below shows that local serving is active.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-bash "><!-- HTML_TAG_START -->$ hermes memory status
<span class="hljs-comment"># Memory status</span>
<span class="hljs-comment">#────────────────────────────────────────</span>
<span class="hljs-comment"># Built-in: always active</span>
<span class="hljs-comment"># Provider: (none — built-in only)</span><!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="opencode" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#opencode"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>OpenCode</span></h2> <p data-svelte-h="svelte-8mfop6"><a href="https://opencode.ai" rel="nofollow">OpenCode</a> works locally with llama.cpp. Define a <code>~/.config/opencode/opencode.json</code>:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-json "><!-- HTML_TAG_START --><span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;$schema&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-string">&quot;https://opencode.ai/config.json&quot;</span><span class="hljs-punctuation">,</span>
 <span class="hljs-attr">&quot;provider&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;llama.cpp&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;npm&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-string">&quot;@ai-sdk/openai-compatible&quot;</span><span class="hljs-punctuation">,</span>
 <span class="hljs-attr">&quot;name&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-string">&quot;llama-server (local)&quot;</span><span class="hljs-punctuation">,</span>
 <span class="hljs-attr">&quot;options&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;baseURL&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-string">&quot;http://127.0.0.1:8080/v1&quot;</span>
 <span class="hljs-punctuation">}</span><span class="hljs-punctuation">,</span>
 <span class="hljs-attr">&quot;models&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;gemma-4-26b-4b-it&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;name&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-string">&quot;Gemma 4 (local)&quot;</span><span class="hljs-punctuation">,</span>
 <span class="hljs-attr">&quot;limit&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-punctuation">{</span>
 <span class="hljs-attr">&quot;context&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-number">128000</span><span class="hljs-punctuation">,</span>
 <span class="hljs-attr">&quot;output&quot;</span><span class="hljs-punctuation">:</span> <span class="hljs-number">8192</span>
 <span class="hljs-punctuation">}</span>
 <span class="hljs-punctuation">}</span>
 <span class="hljs-punctuation">}</span>
 <span class="hljs-punctuation">}</span>
 <span class="hljs-punctuation">}</span>
<span class="hljs-punctuation">}</span><!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="how-it-works" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#how-it-works"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>How It Works</span></h2> <p data-svelte-h="svelte-lyrti7">The setup has two components running locally:</p> <ol data-svelte-h="svelte-1r3ty1a"><li><strong>llama.cpp server</strong> — Serves the model as an OpenAI-compatible API on <code>localhost</code>.</li> <li><strong>Your agent</strong> — The agent process that sends prompts to the local server, reasons about tasks, and executes actions.</li></ol> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=" "><!-- HTML_TAG_START -->┌─────────┐ API calls ┌──────────────────┐
│ Agent │ ───────────────▶ │ llama.cpp server │
│ │ ◀─────────────── │ (local <span class="hljs-keyword">model</span>) │
└─────────┘ responses └──────────────────┘
 │
 ▼
 Your <span class="hljs-keyword">files</span>,
 terminal, etc.<!-- HTML_TAG_END --></pre></div> <h2 class="relative group"><a id="alternative-llama-agent" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#alternative-llama-agent"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Alternative: llama-agent</span></h2> <p data-svelte-h="svelte-xrdnfz"><a href="https://github.com/gary149/llama-agent" rel="nofollow">llama-agent</a> takes a different approach — it builds the agent loop directly into <a href="https://github.com/ggerganov/llama.cpp" rel="nofollow">llama.cpp</a> as a single binary with zero external dependencies. No Node.js, no Python, just compile and run:</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class="language-bash "><!-- HTML_TAG_START -->git <span class="hljs-built_in">clone</span> https://github.com/gary149/llama-agent.git
<span class="hljs-built_in">cd</span> llama-agent
<span class="hljs-comment"># Build</span>
cmake -B build
cmake --build build --target llama-agent
<span class="hljs-comment"># Run (downloads the model automatically)</span>
./build/bin/llama-agent -hf ggml-org/gemma-4-26b-a4b-it-GGUF:Q4_K_M<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1cxz41v">Because tool calls happen in-process rather than over HTTP, there is no network overhead between the model and the agent. It also supports subagents, MCP servers, and an HTTP API server mode.</p> <h2 class="relative group"><a id="next-steps" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#next-steps"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Next Steps</span></h2> <ul data-svelte-h="svelte-azmnuf"><li><a href="./local-apps">Use AI Models Locally</a> — Learn more about running models on your machine</li> <li><a href="./gguf-llamacpp">llama.cpp Guide</a> — Detailed llama.cpp installation and usage</li> <li><a href="./agents-overview">Agents on the Hub</a> — Connect agents to the Hugging Face ecosystem</li></ul> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/hub-docs/blob/main/docs/hub/agents-local.md" target="_blank"><svg class="mr-1" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M31,16l-7,7l-1.41-1.41L28.17,16l-5.58-5.59L24,9l7,7z"></path><path d="M1,16l7-7l1.41,1.41L3.83,16l5.58,5.59L8,23l-7-7z"></path><path d="M12.419,25.484L17.639,6.552l1.932,0.518L14.351,26.002z"></path></svg> <span data-svelte-h="svelte-zjs2n5"><span class="underline">Update</span> on GitHub</span></a> <p></p>
<script>
 {
 __sveltekit_16dt96c = {
 assets: "/docs/hub/pr_2602/en",
 base: "/docs/hub/pr_2602/en",
 env: {}
 };
 const element = document.currentScript.parentElement;
 const data = [null,null];
 Promise.all([
 import("/docs/hub/pr_2602/en/_app/immutable/entry/start.cc882a17.js"),
 import("/docs/hub/pr_2602/en/_app/immutable/entry/app.07a28e47.js")
 ]).then(([kit, app]) => {
 kit.start(app, element, {
 node_ids: [0, 9],
 data,
 form: null,
 error: null
 });
 });
 }
 </script>

Xet Storage Details

Size:
58.4 kB
·
Xet hash:
a377ee3b5745941d6e72f1e7c15e3f0c3cd6d5a49d758b79e5936d3e4297af2c

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.