PageSourceSearch

https://www.xybrid.ai/models

html xybrid.ai collected 2026-10-05 13:23:41 UTC 123,365 bytes, 57 lines download raw bytes

1<!doctype html>
2<html lang="en">
3	<head>
4		<meta charset="utf-8" />
5		<meta name="viewport" content="width=device-width, initial-scale=1" />
6		<link href="https://fonts.googleapis.com/css2?family=Tilt+Neon&display=swap" rel="stylesheet" />
7		<link rel="apple-touch-icon" sizes="57x57" href="/favicon-57x57.png">
8		<link rel="apple-touch-icon" sizes="60x60" href="/favicon-60x60.png">
9		<link rel="apple-touch-icon" sizes="72x72" href="/favicon-72x72.png">
10		<link rel="apple-touch-icon" sizes="76x76" href="/favicon-76x76.png">
11		<link rel="apple-touch-icon" sizes="114x114" href="/favicon-114x114.png">
12		<link rel="apple-touch-icon" sizes="120x120" href="/favicon-120x120.png">
13		<link rel="apple-touch-icon" sizes="144x144" href="/favicon-144x144.png">
14		<link rel="apple-touch-icon" sizes="152x152" href="/favicon-152x152.png">
15		<link rel="apple-touch-icon" sizes="180x180" href="/favicon-180x180.png">
16		<link rel="icon" type="image/png" sizes="16x16" href="/favicon-16x16.png">
17		<link rel="icon" type="image/png" sizes="32x32" href="/favicon-32x32.png">
18		<link rel="icon" type="image/png" sizes="96x96" href="/favicon-96x96.png">
19		<link rel="icon" type="image/png" sizes="192x192" href="/favicon-192x192.png">
20		<link rel="shortcut icon" type="image/x-icon" href="/favicon.ico">
21		<link rel="icon" type="image/x-icon" href="/favicon.ico">
22		<meta name="msapplication-TileColor" content="#ffffff">
23		<meta name="msapplication-TileImage" content="/favicon-144x144.png">
24		<meta name="msapplication-config" content="/browserconfig.xml">
25		
26		<link href="./_app/immutable/assets/0.DG2aLNwg.css" rel="stylesheet">
27		<link href="./_app/immutable/assets/Footer.BwIXAKsC.css" rel="stylesheet">
28		<link href="./_app/immutable/assets/8.CkYyeexZ.css" rel="stylesheet"><!--12qhfyh--><meta name="description" content="Browse 37 on-device AI models for text generation, text-to-speech, and speech recognition. Run locally on iOS, Android, macOS, and more."/> <meta property="og:type" content="website"/> <meta property="og:url" content="https://www.xybrid.ai/models"/> <meta property="og:title" content="Model Registry | Xybrid"/> <meta property="og:description" content="Browse 37 on-device AI models for text generation, text-to-speech, and speech recognition. Run locally on iOS, Android, macOS, and more."/> <meta property="og:image" content="https://www.xybrid.ai/og-image.jpg"/> <meta property="og:image:width" content="1200"/> <meta property="og:image:height" content="630"/> <meta property="og:site_name" content="Xybrid"/> <meta property="og:logo" content="https://www.xybrid.ai/logo.png"/> <meta name="twitter:card" content="summary_large_image"/> <meta name="twitter:url" content="https://www.xybrid.ai/models"/> <meta name="twitter:title" content="Model Registry | Xybrid"/> <meta name="twitter:description" content="Browse 37 on-device AI models for text generation, text-to-speech, and speech recognition. Run locally on iOS, Android, macOS, and more."/> <meta name="twitter:image" content="https://www.xybrid.ai/og-image.jpg"/> <link rel="canonical" href="https://www.xybrid.ai/models"/> <link rel="alternate" type="application/rss+xml" title="Xybrid Blog" href="/blog/rss.xml"/><!----><title>Model Registry | Xybrid</title>
29	</head>
30	<body data-sveltekit-preload-data="hover">
31		<div style="display: contents"><!--[--><!--[--><!----><!----><div class="x-page x-root svelte-18pldtr"><nav class="x-nav "><div class="x-nav__left"><a class="x-nav__brand" href="/" aria-label="Xybrid home"><img class="x-logo-mark" src="/logo.png" alt="" width="34" height="34" style="display: inline-block; width: 34px; height: 34px; object-fit: contain;"/><!----></a> <ul class="x-nav__links"><li><a class="x-nav__link" href="/models">Models</a></li> <li><a class="x-nav__link" href="/blog">Blog</a></li> <li><a class="x-nav__link" href="/games">Games</a></li> <li><a class="x-nav__link --has-arrow" href="https://docs.xybrid.dev" target="_blank" rel="noopener">Docs</a></li></ul></div> <div class="x-nav__right"><a class="x-btn --ghost --sm x-nav__social svelte-1h8gbq9" href="https://github.com/xybrid-ai/xybrid" target="_blank" rel="noopener"><svg aria-hidden="true" width="14" height="14" viewBox="0 0 16 16" fill="currentColor"><!--[!--><!--[--><path d="M8 .2A8 8 0 0 0 .5 11a8 8 0 0 0 5.5 5.06c.4.07.55-.18.55-.4v-1.4c-2.24.5-2.7-1.07-2.7-1.07-.36-.94-.9-1.18-.9-1.18-.74-.5.06-.5.06-.5.82.06 1.25.85 1.25.85.73 1.26 1.92.9 2.4.68.07-.53.28-.9.5-1.1-1.78-.2-3.65-.9-3.65-4 0-.88.32-1.6.84-2.17-.08-.2-.36-1.02.08-2.12 0 0 .68-.22 2.23.83a7.7 7.7 0 0 1 4.05 0c1.55-1.05 2.23-.83 2.23-.83.44 1.1.16 1.92.08 2.12.52.57.84 1.3.84 2.17 0 3.1-1.88 3.8-3.67 4 .3.25.55.74.55 1.5v2.22c0 .22.15.47.55.4A8 8 0 0 0 15.5 11 8 8 0 0 0 8 .2z"></path><!--]--><!--]--></svg><!----><!--[-->465<!--]--></a> <a class="x-btn --ghost --sm x-nav__social svelte-1h8gbq9" href="https://discord.gg/xybrid" target="_blank" rel="noopener"><svg aria-hidden="true" width="14" height="14" viewBox="0 0 16 16" fill="currentColor"><!--[!--><!--[!--><!--[--><path d="M13.55 3.06A12.4 12.4 0 0 0 10.5 2.13a8.5 8.5 0 0 0-.4.83 11.5 11.5 0 0 0-3.43 0 8.4 8.4 0 0 0-.4-.83 12.4 12.4 0 0 0-3.05.93A12.93 12.93 0 0 0 .42 11.6a12.5 12.5 0 0 0 3.78 1.9c.3-.42.58-.86.81-1.32a8.1 8.1 0 0 1-1.28-.62c.1-.08.21-.16.32-.24a8.93 8.93 0 0 0 7.5 0c.11.08.22.16.32.24-.41.24-.83.45-1.28.62.24.46.5.9.81 1.32a12.5 12.5 0 0 0 3.78-1.9 12.78 12.78 0 0 0-2.63-8.54ZM5.35 9.98c-.74 0-1.35-.69-1.35-1.54s.6-1.55 1.35-1.55c.74 0 1.36.7 1.35 1.55 0 .85-.6 1.54-1.35 1.54Zm5.3 0c-.74 0-1.35-.69-1.35-1.54s.6-1.55 1.35-1.55c.74 0 1.36.7 1.35 1.55 0 .85-.6 1.54-1.35 1.54Z"></path><!--]--><!--]--><!--]--></svg><!----> Discord</a> <a class="x-btn --ghost --sm x-nav__signin svelte-1h8gbq9" href="https://console.xybrid.dev">Sign in</a> <a class="x-btn --primary --sm" href="https://console.xybrid.dev/">Get API key <svg class="x-arrow" aria-hidden="true" width="12" height="12" viewBox="0 0 13 13" fill="none"><!--[--><path d="M3 6.5h6.5M7 4l2.5 2.5L7 9" stroke="currentColor" stroke-width="1.4" stroke-linecap="round" stroke-linejoin="round"></path><!--]--></svg><!----></a> <span class="x-nav__progress" aria-hidden="true"></span> <button type="button" class="x-nav__burger svelte-1h8gbq9" aria-label="Open menu" aria-expanded="false" aria-controls="x-nav-drawer"><svg aria-hidden="true" width="18" height="18" viewBox="0 0 18 18" fill="none"><!--[!--><!--[!--><!--[!--><!--[--><path d="M3 5h12M3 9h12M3 13h12" stroke="currentColor" stroke-width="1.5" stroke-linecap="round"></path><!--]--><!--]--><!--]--><!--]--></svg><!----></button></div></nav> <!--[!--><!--]--> <div id="x-nav-drawer" class="x-nav__drawer svelte-1h8gbq9" aria-hidden="true" inert>
31<ul class="x-nav__drawer-links svelte-1h8gbq9"><li class="svelte-1h8gbq9"><a href="/models" class="svelte-1h8gbq9">Models</a></li> <li class="svelte-1h8gbq9"><a href="/blog" class="svelte-1h8gbq9">Blog</a></li> <li class="svelte-1h8gbq9"><a href="/games" class="svelte-1h8gbq9">Games</a></li> <li class="svelte-1h8gbq9"><a href="https://docs.xybrid.dev" target="_blank" rel="noopener" class="svelte-1h8gbq9">Docs ↗</a></li></ul> <div class="x-nav__drawer-foot svelte-1h8gbq9"><a class="x-btn --ghost" href="https://github.com/xybrid-ai/xybrid" target="_blank" rel="noopener"><svg aria-hidden="true" width="14" height="14" viewBox="0 0 16 16" fill="currentColor"><!--[!--><!--[--><path d="M8 .2A8 8 0 0 0 .5 11a8 8 0 0 0 5.5 5.06c.4.07.55-.18.55-.4v-1.4c-2.24.5-2.7-1.07-2.7-1.07-.36-.94-.9-1.18-.9-1.18-.74-.5.06-.5.06-.5.82.06 1.25.85 1.25.85.73 1.26 1.92.9 2.4.68.07-.53.28-.9.5-1.1-1.78-.2-3.65-.9-3.65-4 0-.88.32-1.6.84-2.17-.08-.2-.36-1.02.08-2.12 0 0 .68-.22 2.23.83a7.7 7.7 0 0 1 4.05 0c1.55-1.05 2.23-.83 2.23-.83.44 1.1.16 1.92.08 2.12.52.57.84 1.3.84 2.17 0 3.1-1.88 3.8-3.67 4 .3.25.55.74.55 1.5v2.22c0 .22.15.47.55.4A8 8 0 0 0 15.5 11 8 8 0 0 0 8 .2z"></path><!--]--><!--]--></svg><!----><!--[-->465<!--]--></a> <a class="x-btn --ghost" href="https://discord.gg/xybrid" target="_blank" rel="noopener"><svg aria-hidden="true" width="14" height="14" viewBox="0 0 16 16" fill="currentColor"><!--[!--><!--[!--><!--[--><path d="M13.55 3.06A12.4 12.4 0 0 0 10.5 2.13a8.5 8.5 0 0 0-.4.83 11.5 11.5 0 0 0-3.43 0 8.4 8.4 0 0 0-.4-.83 12.4 12.4 0 0 0-3.05.93A12.93 12.93 0 0 0 .42 11.6a12.5 12.5 0 0 0 3.78 1.9c.3-.42.58-.86.81-1.32a8.1 8.1 0 0 1-1.28-.62c.1-.08.21-.16.32-.24a8.93 8.93 0 0 0 7.5 0c.11.08.22.16.32.24-.41.24-.83.45-1.28.62.24.46.5.9.81 1.32a12.5 12.5 0 0 0 3.78-1.9 12.78 12.78 0 0 0-2.63-8.54ZM5.35 9.98c-.74 0-1.35-.69-1.35-1.54s.6-1.55 1.35-1.55c.74 0 1.36.7 1.35 1.55 0 .85-.6 1.54-1.35 1.54Zm5.3 0c-.74 0-1.35-.69-1.35-1.54s.6-1.55 1.35-1.55c.74 0 1.36.7 1.35 1.55 0 .85-.6 1.54-1.35 1.54Z"></path><!--]--><!--]--><!--]--></svg><!----> Discord</a> <a class="x-btn --ghost" href="https://console.xybrid.dev">Sign in</a></div></div><!----> <main><section class="x-section"><div class="x-container"><div class="x-section__head"><span class="x-eyebrow">Registry</span> <h1 class="x-h1" style="max-width: 20ch;">37 models. <em>Ready to run.</em></h1> <p class="x-lead">On-device, cross-platform. Browse the catalog, copy the integration snippet, ship.</p></div> <div class="x-models__toolbar svelte-18pldtr"><label class="x-models__search svelte-18pldtr"><svg width="16" height="16" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.6" aria-hidden="true"><circle cx="11" cy="11" r="7"></circle><path d="m20 20-3.5-3.5"></path></svg> <input type="search" placeholder="Search models, families, descriptions…" value="" class="svelte-18pldtr"/></label> <div class="x-models__filters svelte-18pldtr"><div class="x-models__group svelte-18pldtr" role="group" aria-label="Task"><!--[--><button type="button" class="x-chip svelte-18pldtr x-chip--active">All tasks</button><button type="button" class="x-chip svelte-18pldtr">Text generation</button><button type="button" class="x-chip svelte-18pldtr">Text → Speech</button><button type="button" class="x-chip svelte-18pldtr">Speech → Text</button><!--]--></div> <div class="x-models__group svelte-18pldtr" role="group" aria-label="Size"><!--[--><button type="button" class="x-chip x-chip--sm svelte-18pldtr x-chip--active">Any size</button><button type="button" class="x-chip x-chip--sm svelte-18pldtr">&lt; 100M</button><button type="button" class="x-chip x-chip--sm svelte-18pldtr">100M – 1B</button><button type="button" class="x-chip x-chip--sm svelte-18pldtr">1B+</button><button type="button" class="x-chip x-chip--sm svelte-18pldtr">7B+</button><!--]--></div></div></div> <p class="x-models__count svelte-18pldtr">Showing 37 of 37</p> <!--[--><div class="x-models__grid svelte-18pldtr"><!--[--><a class="x-model-card svelte-18pldtr" href="/models/bonsai-27b"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-fg-3); border-color: color-mix(in srgb, var(--c-fg-3) 32%, var(--c-line));">vision-language</span> <span class="x-model-card__format svelte-18pldtr">
31gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[!--><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Bonsai 27B</h3></div> <p class="x-model-card__desc svelte-18pldtr">PrismML Bonsai 27B 1-bit vision-language model for local reasoning.</p> <div class="x-model-card__meta svelte-18pldtr"><span>Prismml</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>27.3B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>3.5 GB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/functiongemma-270m-it"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[!--><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Functiongemma 270M It</h3></div> <p class="x-model-card__desc svelte-18pldtr">Google FunctionGemma 270M instruction-tuned function-calling model for local inference.</p> <div class="x-model-card__meta svelte-18pldtr"><span>Functiongemma</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>270M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>278.1 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/gemma-3-1b"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/gemma.png" alt="Google" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Gemma 3 1B</h3></div> <p class="x-model-card__desc svelte-18pldtr">Gemma 3 1B - Google's mobile-optimized instruction-tuned LLM with 32K context</p> <div class="x-model-card__meta svelte-18pldtr"><span>Google</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>1B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>768.7 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/gemma-4-e2b"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/gemma.png" alt="Google" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Gemma 4 E2b</h3></div> <p class="x-model-card__desc svelte-18pldtr">Gemma 4 E2B - Google's compact multimodal LLM with 2.3B effective params, 128K context, and audio/image/video understanding</p> <div class="x-model-card__meta svelte-18pldtr"><span>Google</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>5.1B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>2.9 GB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/gemma-4-e4b"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/gemma.png" alt="Google" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Gemma 4 E4b</h3></div> <p class="x-model-card__desc svelte-18pldtr">Gemma 4 E4B - Google's mid-range multimodal LLM with 4.5B effective params, 128K context, and audio/image/video understanding</p> <div class="x-model-card__meta svelte-18pldtr"><span>Google</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>8B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>4.6 GB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/gemma3npc-1b"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">
31gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/gemma.png" alt="Google" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Gemma3npc 1B</h3></div> <p class="x-model-card__desc svelte-18pldtr">Gemma3NPC 1B - Mobile-friendly NPC roleplay fine-tune of Gemma 3 1B for in-character game dialogue</p> <div class="x-model-card__meta svelte-18pldtr"><span>Google</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>1B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>768.7 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/gemma3npc-it"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/gemma.png" alt="Google" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Gemma3npc It</h3></div> <p class="x-model-card__desc svelte-18pldtr">Gemma3NPC-it - NPC roleplay fine-tune of Gemma3n-E4B for in-character game dialogue</p> <div class="x-model-card__meta svelte-18pldtr"><span>Google</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>7B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>3.9 GB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/kitten-tts-micro-0.8"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-mint); border-color: color-mix(in srgb, var(--c-d-mint) 32%, var(--c-line));">Text to Speech</span> <span class="x-model-card__format svelte-18pldtr">onnx</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/kittenml.svg" alt="KittenML" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Kitten Tts Micro 0.8</h3></div> <p class="x-model-card__desc svelte-18pldtr">KittenTTS Micro 0.8 - Compact StyleTTS 2 (40M params, 8 named voices, OpenPhonemizer)</p> <div class="x-model-card__meta svelte-18pldtr"><span>KittenML</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>40M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>90.2 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/kitten-tts-mini-0.8"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-mint); border-color: color-mix(in srgb, var(--c-d-mint) 32%, var(--c-line));">Text to Speech</span> <span class="x-model-card__format svelte-18pldtr">onnx</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/kittenml.svg" alt="KittenML" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Kitten Tts Mini 0.8</h3></div> <p class="x-model-card__desc svelte-18pldtr">KittenTTS Mini 0.8 - High-quality StyleTTS 2 (80M params, 8 named voices, OpenPhonemizer)</p> <div class="x-model-card__meta svelte-18pldtr"><span>KittenML</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>80M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>111.1 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/kitten-tts-nano-0.2"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-mint); border-color: color-mix(in srgb, var(--c-d-mint) 32%, var(--c-line));">Text to Speech</span> <span class="x-model-card__format svelte-18pldtr">onnx</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/kittenml.svg" alt="KittenML" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Kitten Tts Nano 0.2</h3></div> <p class="x-model-card__desc svelte-18pldtr">KittenTTS Nano 0.2 - Ultra-lightweight TTS (15M params, &lt;25MB)</p> <div class="x-model-card__meta svelte-18pldtr"><span>KittenML</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>15M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>18.3 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/kitten-tts-nano-0.8"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-mint); border-color: color-mix(in srgb, var(--c-d-mint) 32%, var(--c-line));">Text to Speech</span> <span class="x-model-card__format svelte-18pldtr">onnx</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/kittenml.svg" alt="KittenML" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Kitten Tts Nano 0.8</h3></div> <p class="x-model-card__desc svelte-18pldtr">KittenTTS Nano 0.8 - Ultra-lightweight StyleTTS 2 (15M params, 8 named voices, OpenPhonemizer)</p> <div class="x-model-card__meta svelte-18pldtr"><span>KittenML</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>15M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>78.3 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/kokoro-82m"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-mint); border-color: color-mix(in srgb, var(--c-d-mint) 32%, var(--c-line));">Text to Speech</span> <span class="x-model-card__format svelte-18pldtr">onnx</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[!--><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Kokoro 82M</h3></div> <p class="x-model-card__desc svelte-18pldtr">Kokoro 82M - High-quality TTS with 24 voices, Misaki dictionary</p> <div class="x-model-card__meta svelte-18pldtr"><span>Hexgrad</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>82M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>174.9 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/lfm2-vl-450m"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-fg-3); border-color: color-mix(in srgb, var(--c-fg-3) 32%, var(--c-line));">vision-language</span> <span class="x-model-card__format svelte-18pldtr">
31gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/liquid-ai.svg" alt="Liquid AI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Lfm2 Vl 450M</h3></div> <p class="x-model-card__desc svelte-18pldtr">Liquid LFM2-VL 450M compact vision-language model for edge inference.</p> <div class="x-model-card__meta svelte-18pldtr"><span>Liquid AI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>450M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>209.1 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/lfm2.5-1.2b-instruct"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/liquid-ai.svg" alt="Liquid AI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Lfm2.5 1.2B Instruct</h3></div> <p class="x-model-card__desc svelte-18pldtr">Liquid LFM2.5 1.2B Instruct - hybrid conv+attention LLM optimized for edge deployment (agentic tasks, data extraction, RAG, tool calling)</p> <div class="x-model-card__meta svelte-18pldtr"><span>Liquid AI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>1.2B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>697.0 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/lfm2.5-1.2b-thinking"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/liquid-ai.svg" alt="Liquid AI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Lfm2.5 1.2B Thinking</h3></div> <p class="x-model-card__desc svelte-18pldtr">Liquid LFM2.5 1.2B Thinking - hybrid conv+attention reasoning LLM optimized for edge deployment (fast CPU inference, low memory)</p> <div class="x-model-card__meta svelte-18pldtr"><span>Liquid AI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>1.2B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>697.0 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/lfm2.5-230m"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/liquid-ai.svg" alt="Liquid AI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Lfm2.5 230M</h3></div> <p class="x-model-card__desc svelte-18pldtr">Liquid LFM2.5 230M - smallest hybrid conv+attention LLM optimized for edge deployment (9 languages, tool calling, fast CPU inference)</p> <div class="x-model-card__meta svelte-18pldtr"><span>Liquid AI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>230M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>146.3 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/lfm2.5-350m"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">
31gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/liquid-ai.svg" alt="Liquid AI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Lfm2.5 350M</h3></div> <p class="x-model-card__desc svelte-18pldtr">Liquid LFM2.5 350M - hybrid conv+attention LLM optimized for edge deployment (9 languages, tool calling)</p> <div class="x-model-card__meta svelte-18pldtr"><span>Liquid AI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>354M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>218.7 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/lfm2.5-vl-3b"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-fg-3); border-color: color-mix(in srgb, var(--c-fg-3) 32%, var(--c-line));">vision-language</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/liquid-ai.svg" alt="Liquid AI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Lfm2.5 Vl 3B</h3></div> <p class="x-model-card__desc svelte-18pldtr">Liquid LFM2.5-VL 3B vision-language model for local inference.</p> <div class="x-model-card__meta svelte-18pldtr"><span>Liquid AI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>3B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>1.6 GB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/llama-3.2-1b"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/meta.png" alt="Meta" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Llama 3.2 1B</h3></div> <p class="x-model-card__desc svelte-18pldtr">Llama 3.2 1B - Meta's lightweight mobile-optimized instruction-tuned LLM with 128K context</p> <div class="x-model-card__meta svelte-18pldtr"><span>Meta</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>1B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>770.3 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/ministral-3-3b"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/mistral.png" alt="Mistral" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Ministral 3 3B</h3></div> <p class="x-model-card__desc svelte-18pldtr">Ministral 3 3B Instruct - Mistral AI's edge-optimized instruction-tuned LLM with 256K context</p> <div class="x-model-card__meta svelte-18pldtr"><span>Mistral</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>3.4B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>2.0 GB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/mistral-7b"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/mistral.png" alt="Mistral" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Mistral 7B</h3></div> <p class="x-model-card__desc svelte-18pldtr">Mistral 7B Instruct v0.3 - High-quality desktop LLM with function calling support</p> <div class="x-model-card__meta svelte-18pldtr"><span>Mistral</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>7B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>4.1 GB</span></div></a>
31<a class="x-model-card svelte-18pldtr" href="/models/neutts-air-q4"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-mint); border-color: color-mix(in srgb, var(--c-d-mint) 32%, var(--c-line));">Text to Speech</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/neuphonic.png" alt="Neuphonic" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Neutts Air Q4</h3></div> <p class="x-model-card__desc svelte-18pldtr">NeuTTS Air (~500M) — codec TTS with voice cloning. Q4 GGUF backbone + NeuCodec decoder. Higher quality than Nano.</p> <div class="x-model-card__meta svelte-18pldtr"><span>Neuphonic</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>500M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>755.0 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/neutts-nano-q4"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-mint); border-color: color-mix(in srgb, var(--c-d-mint) 32%, var(--c-line));">Text to Speech</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/neuphonic.png" alt="Neuphonic" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Neutts Nano Q4</h3></div> <p class="x-model-card__desc svelte-18pldtr">NeuTTS Nano (120M) — codec TTS with voice cloning. Q4 GGUF backbone + NeuCodec ONNX decoder.</p> <div class="x-model-card__meta svelte-18pldtr"><span>Neuphonic</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>120M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>469.7 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/phi4-mini"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/microsoft.png" alt="Microsoft" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Phi4 Mini</h3></div> <p class="x-model-card__desc svelte-18pldtr">Phi-4 Mini 3.8B - Microsoft's compact reasoning LLM with 128K context</p> <div class="x-model-card__meta svelte-18pldtr"><span>Microsoft</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>3.8B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>2.3 GB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/qwen2.5-0.5b-instruct"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/qwen.png" alt="Qwen" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Qwen2.5 0.5B Instruct</h3></div> <p class="x-model-card__desc svelte-18pldtr">Qwen 2.5 0.5B Instruct - Small but capable instruction-tuned LLM from Alibaba Cloud</p> <div class="x-model-card__meta svelte-18pldtr"><span>Qwen</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>500M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>468.6 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/qwen3.5-0.8b"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">
31gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/qwen.png" alt="Qwen" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Qwen3.5 0.8B</h3></div> <p class="x-model-card__desc svelte-18pldtr">Qwen 3.5 0.8B - Alibaba Cloud's lightweight multimodal LLM (text-only mode, 201 languages)</p> <div class="x-model-card__meta svelte-18pldtr"><span>Qwen</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>800M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>495.7 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/qwen3.5-2b"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/qwen.png" alt="Qwen" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Qwen3.5 2B</h3></div> <p class="x-model-card__desc svelte-18pldtr">Qwen 3.5 2B - Alibaba Cloud's compact multimodal LLM (text-only mode, 201 languages)</p> <div class="x-model-card__meta svelte-18pldtr"><span>Qwen</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>2B params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>1.2 GB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/smollm2-360m"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-accent-2); border-color: color-mix(in srgb, var(--c-accent-2) 32%, var(--c-line));">Text Generation</span> <span class="x-model-card__format svelte-18pldtr">gguf</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/huggingface.png" alt="HuggingFace" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Smollm2 360M</h3></div> <p class="x-model-card__desc svelte-18pldtr">SmolLM2 360M - HuggingFace's best tiny LLM, excellent quality/size ratio</p> <div class="x-model-card__meta svelte-18pldtr"><span>HuggingFace</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>360M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>258.1 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/wav2vec2-base-960h"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-peach); border-color: color-mix(in srgb, var(--c-d-peach) 32%, var(--c-line));">Speech Recognition</span> <span class="x-model-card__format svelte-18pldtr">onnx</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/meta.png" alt="Meta" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Wav2vec2 Base 960H</h3></div> <p class="x-model-card__desc svelte-18pldtr">Wav2Vec2 Base 960h - English ASR with CTC decoding</p> <div class="x-model-card__meta svelte-18pldtr"><span>Meta</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>95M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>220.3 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/whisper-base-en-ggml"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-peach); border-color: color-mix(in srgb, var(--c-d-peach) 32%, var(--c-line));">Speech Recognition</span> <span class="x-model-card__format svelte-18pldtr">ggml</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/openai.png" alt="OpenAI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Whisper Base En Ggml</h3></div> <p class="x-model-card__desc svelte-18pldtr">Whisper Base English - English-only ASR (whisper.cpp/GGML runtime)</p> <div class="x-model-card__meta svelte-18pldtr"><span>OpenAI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>74M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>57.0 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/whisper-base-ggml"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-peach); border-color: color-mix(in srgb, var(--c-d-peach) 32%, var(--c-line));">Speech Recognition</span> <span class="x-model-card__format svelte-18pldtr">ggml</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/openai.png" alt="OpenAI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Whisper Base Ggml</h3></div> <p class="x-model-card__desc svelte-18pldtr">Whisper Base - Multilingual ASR (whisper.cpp/GGML runtime)</p> <div class="x-model-card__meta svelte-18pldtr"><span>OpenAI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>74M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>56.9 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/whisper-large-v3-turbo-ggml"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-peach); border-color: color-mix(in srgb, var(--c-d-peach) 32%, var(--c-line));">Speech Recognition</span> <span class="x-model-card__format svelte-18pldtr">ggml</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/openai.png" alt="OpenAI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Whisper Large V3 Turbo Ggml</h3></div> <p class="x-model-card__desc svelte-18pldtr">Whisper Large v3 Turbo - Highest-accuracy multilingual ASR (whisper.cpp/GGML runtime)</p> <div class="x-model-card__meta svelte-18pldtr"><span>OpenAI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>809M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>547.4 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/whisper-medium-ggml"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-peach); border-color: color-mix(in srgb, var(--c-d-peach) 32%, var(--c-line));">Speech Recognition</span> <span class="x-model-card__format svelte-18pldtr">ggml</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/openai.png" alt="OpenAI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Whisper Medium Ggml</h3></div> <p class="x-model-card__desc svelte-18pldtr">Whisper Medium - Multilingual ASR (whisper.cpp/GGML runtime)</p> <div class="x-model-card__meta svelte-18pldtr"><span>OpenAI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>769M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>514.2 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/whisper-small-ggml"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-peach); border-color: color-mix(in srgb, var(--c-d-peach) 32%, var(--c-line));">Speech Recognition</span> <span class="x-model-card__format svelte-18pldtr">ggml</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/openai.png" alt="OpenAI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Whisper Small Ggml</h3></div> <p class="x-model-card__desc svelte-18pldtr">Whisper Small - Multilingual ASR (whisper.cpp/GGML runtime)</p> <div class="x-model-card__meta svelte-18pldtr"><span>OpenAI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>244M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>181.3 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/whisper-tiny"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-peach); border-color: color-mix(in srgb, var(--c-d-peach) 32%, var(--c-line));">Speech Recognition</span> <span class="x-model-card__format svelte-18pldtr">safetensors</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/openai.png" alt="OpenAI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Whisper Tiny</h3></div> <p class="x-model-card__desc svelte-18pldtr">Whisper Tiny - Fast multilingual ASR (Candle/SafeTensors runtime)</p> <div class="x-model-card__meta svelte-18pldtr"><span>OpenAI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>39M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>84.7 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/whisper-tiny-en-ggml"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-peach); border-color: color-mix(in srgb, var(--c-d-peach) 32%, var(--c-line));">Speech Recognition</span> <span class="x-model-card__format svelte-18pldtr">ggml</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/openai.png" alt="OpenAI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Whisper Tiny En Ggml</h3></div> <p class="x-model-card__desc svelte-18pldtr">Whisper Tiny English - Fastest English-only ASR (whisper.cpp/GGML runtime)</p> <div class="x-model-card__meta svelte-18pldtr"><span>OpenAI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>39M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>30.7 MB</span></div></a><a class="x-model-card svelte-18pldtr" href="/models/whisper-tiny-ggml"><div class="x-model-card__head svelte-18pldtr"><span class="x-model-card__task svelte-18pldtr" style="color: var(--c-d-peach); border-color: color-mix(in srgb, var(--c-d-peach) 32%, var(--c-line));">Speech Recognition</span> <span class="x-model-card__format svelte-18pldtr">ggml</span></div> <div class="x-model-card__title-row svelte-18pldtr"><!--[--><span class="x-model-card__logo svelte-18pldtr"><img src="/logos/ai/openai.png" alt="OpenAI" loading="lazy" class="svelte-18pldtr"/></span><!--]--> <h3 class="x-model-card__title svelte-18pldtr">Whisper Tiny Ggml</h3></div> <p class="x-model-card__desc svelte-18pldtr">Whisper Tiny - Fast multilingual ASR (whisper.cpp/GGML runtime)</p> <div class="x-model-card__meta svelte-18pldtr"><span>OpenAI</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>39M params</span> <span class="x-model-card__sep svelte-18pldtr">·</span> <span>30.7 MB</span></div></a>
31<!--]--></div><!--]--></div></section></main> <footer class="x-footer"><div class="x-container"><div class="x-footer__row"><div class="x-footer__col"><a href="/" aria-label="Xybrid home" style="display: inline-flex; align-items: center; gap: 10px; font-size: 16px; color: var(--c-fg);"><img class="x-logo-mark" src="/logo.png" alt="" width="26" height="26" style="display: inline-block; width: 26px; height: 26px; object-fit: contain;"/><!----></a> <p class="x-small" style="margin-top: 14px; max-width: 36ch;">The operating layer for on-device AI. Built in 🇫🇷.</p></div> <div class="x-footer__col"><h4>Product</h4> <ul><li><a href="https://github.com/xybrid-ai/xybrid" target="_blank" rel="noopener">Runtime ↗</a></li> <li><a href="https://console.xybrid.dev">Console</a></li></ul></div> <div class="x-footer__col"><h4>Developers</h4> <ul><li><a href="https://docs.xybrid.dev" target="_blank" rel="noopener">Documentation ↗</a></li> <li><a href="https://docs.xybrid.dev/quickstart" target="_blank" rel="noopener">Quickstart ↗</a></li> <li><a href="https://docs.xybrid.dev/api">API reference</a></li></ul></div> <div class="x-footer__col"><h4>Company</h4> <ul><li><a href="/blog">Blog</a></li></ul></div></div> <div class="x-footer__bottom"><span>© 2026 Xybrid</span> <span style="font-family: var(--c-font-mono); font-size: 11px; color: var(--c-fg-4); letter-spacing: 0.05em;">v0.10.1</span></div></div></footer><!----></div><!----><!----><!----><!--]--> <!--[!--><!--]--><!--]-->
32			
33			
33<script>
34				{
35					__sveltekit_b1t3p1 = {
36						base: new URL(".", location).pathname.slice(0, -1),
37						env: {"PUBLIC_POSTHOG_KEY":"phc_wChHnifJyQW2PWGRQTTbtdoT9FobMD3rQ2TNfZJAvfcR"}
38					};
39
40					const element = document.currentScript.parentElement;
41
42					Promise.all([
43						import("./_app/immutable/entry/start.DCslHCDY.js"),
44						import("./_app/immutable/entry/app.BFO6RO-y.js")
45					]).then(([kit, app]) => {
46						kit.start(app, element, {
47							node_ids: [0, 8],
48							data: [{type:"data",data:{gh:{stars:465,contributors:15,commitsThisMonth:4,commitGraph:[1,2,2,0,1,2,0,0,0,1,0,0,0,2,2,4,2,4,1,1,5,2,6,3,2,2,2,2,0,2,1,1]},latestReleaseTag:"v0.10.1"},uses:{}},{type:"data",data:{models:[{id:"bonsai-27b",family:"prismml",task:"vision-language",parameters:27300000000,description:"PrismML Bonsai 27B 1-bit vision-language model for local reasoning.",defaultVariant:"macos-arm64",variants:{"macos-arm64":{architecture:"qwen35",artifacts:[{file:"Bonsai-27B-mmproj-Q8_0.gguf",sha256:"eb561d41a7bbeb0fcf04883c8af11078ef6cae0a66862a0b68443cfca495269d",size_bytes:629246880}],backend:"llamacpp",context_length:4096,file:"Bonsai-27B-Q1_0.gguf",format:"gguf",hf_repo:"prism-ml/Bonsai-27B-gguf",max_context_length:262144,parameters:26895998464,passthrough:true,quantization:"q1_0_g128",revision:"f10afb355f104535e3e3e98cf7ab7795c72bd292",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:4096,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Includes the registered vision projector; parameters describe the primary language model only. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/prism-ml/Bonsai-27B-gguf/resolve/f10afb355f104535e3e3e98cf7ab7795c72bd292/Bonsai-27B-Q1_0.gguf","https://huggingface.co/prism-ml/Bonsai-27B-gguf/resolve/f10afb355f104535e3e3e98cf7ab7795c72bd292/Bonsai-27B-mmproj-Q8_0.gguf","https://huggingface.co/prism-ml/Bonsai-27B-gguf/tree/f10afb355f104535e3e3e98cf7ab7795c72bd292","https://huggingface.co/prism-ml/Bonsai-27B-gguf/blob/f10afb355f104535e3e3e98cf7ab7795c72bd292/README.md"],weights_mb:4217},sha256:"17ef842e47450caeb8eaa3ebfbbab5d2f2278b62b79be107985fb69a2f819aa0",size_bytes:3803452480}},voices:void 0,displayName:"Bonsai 27B",parameterLabel:"27.3B",sizeLabel:"3.5 GB",format:"gguf",quantization:"q1_0_g128",hfRepo:"prism-ml/Bonsai-27B-gguf",taskLabel:"vision-language",familyLabel:"Prismml"},{id:"functiongemma-270m-it",family:"functiongemma",task:"text-generation",parameters:270000000,description:"Google FunctionGemma 270M instruction-tuned function-calling model for local inference.",defaultVariant:"universal",variants:{universal:{architecture:"gemma3",backend:"llamacpp",context_length:4096,file:"functiongemma-270m-it-q8_0.gguf",format:"gguf",hf_repo:"ggml-org/functiongemma-270m-it-GGUF",max_context_length:32768,parameters:268098176,passthrough:true,quantization:"q8_0",revision:"2566ce14aedfc14fdd0de955ba67346425e67126",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:4096,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/ggml-org/functiongemma-270m-it-GGUF/resolve/2566ce14aedfc14fdd0de955ba67346425e67126/functiongemma-270m-it-q8_0.gguf","https://huggingface.co/google/functiongemma-270m-it/blob/39eccb091651513a5dfb56892d3714c1b5b8276c/README.md","https://huggingface.co/unsloth/functiongemma-270m-it/blob/79eaaa84125b897f5a0ad6638362b864fec813ad/config.json","https://huggingface.co/ggml-org/functiongemma-270m-it-GGUF/blob/2566ce14aedfc14fdd0de955ba67346425e67126/README.md"],weights_mb:272},sha256:"83940d4dd9676710856f43523bed096164a595a96f6b34771610a03937de5270",size_bytes:291557792}},voices:void 0,displayName:"Functiongemma 270M It",parameterLabel:"270M",sizeLabel:"278.1 MB",format:"gguf",quantization:"q8_0",hfRepo:"ggml-org/functiongemma-270m-it-GGUF",taskLabel:"Text Generation",familyLabel:"Functiongemma"},{id:"gemma-3-1b",family:"google",task:"text-generation",parameters:1000000000,description:"Gemma 3 1B - Google's mobile-optimized instruction-tuned LLM with 32K context",defaultVariant:"universal",variants:{universal:{architecture:"gemma3",backend:"llamacpp",context_length:8192,file:"google_gemma-3-1b-it-Q4_K_M.gguf",format:"gguf",hf_repo:"bartowski/google_gemma-3-1b-it-GGUF",max_context_length:32768,parameters:999885952,passthrough:true,quantization:"q4_k_m",revision:"116f76234503685a98f572982177b11d44ec8ff1",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:8192,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB;
48 actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/bartowski/google_gemma-3-1b-it-GGUF/resolve/116f76234503685a98f572982177b11d44ec8ff1/google_gemma-3-1b-it-Q4_K_M.gguf","https://huggingface.co/google/gemma-3-1b-it/blob/dcc83ea841ab6100d6b47a070329e1ba4cf78752/README.md","https://huggingface.co/unsloth/gemma-3-1b-it/blob/5b11413a10db4e486ef16a20101fd028f8f2499c/config.json"],weights_mb:763},sha256:"12bf0fff8815d5f73a3c9b586bd8fee8e7b248c935de70dec367679873d0f29d",size_bytes:806058496}},voices:void 0,displayName:"Gemma 3 1B",parameterLabel:"1B",sizeLabel:"768.7 MB",format:"gguf",quantization:"q4_k_m",hfRepo:"bartowski/google_gemma-3-1b-it-GGUF",taskLabel:"Text Generation",familyLabel:"Google"},{id:"gemma-4-e2b",family:"google",task:"text-generation",parameters:5100000000,description:"Gemma 4 E2B - Google's compact multimodal LLM with 2.3B effective params, 128K context, and audio/image/video understanding",defaultVariant:"universal",variants:{universal:{architecture:"gemma4",backend:"llamacpp",context_length:8192,file:"gemma-4-E2B-it-Q4_K_M.gguf",format:"gguf",hf_repo:"unsloth/gemma-4-E2B-it-GGUF",max_context_length:131072,parameters:4647450147,passthrough:true,quantization:"q4_k_m",revision:"0314792d7f1f7e229411f620751375812bb9faf2",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:8192,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Includes stored per-layer embeddings, not just the marketed effective parameter count. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/unsloth/gemma-4-E2B-it-GGUF/resolve/0314792d7f1f7e229411f620751375812bb9faf2/gemma-4-E2B-it-Q4_K_M.gguf","https://huggingface.co/google/gemma-4-E2B-it/blob/3e22461f65e89153144f8adb70e3b8c2cc9845a7/config.json","https://huggingface.co/google/gemma-4-E2B-it/blob/3e22461f65e89153144f8adb70e3b8c2cc9845a7/README.md"],weights_mb:2948},sha256:"740185b21d22ceb83a11c3aa62ad5842ef32c70f6096d756bbee85a1e4ec34b8",size_bytes:3106738272}},voices:void 0,displayName:"Gemma 4 E2b",parameterLabel:"5.1B",sizeLabel:"2.9 GB",format:"gguf",quantization:"q4_k_m",hfRepo:"unsloth/gemma-4-E2B-it-GGUF",taskLabel:"Text Generation",familyLabel:"Google"},{id:"gemma-4-e4b",family:"google",task:"text-generation",parameters:8000000000,description:"Gemma 4 E4B - Google's mid-range multimodal LLM with 4.5B effective params, 128K context, and audio/image/video understanding",defaultVariant:"universal",variants:{universal:{architecture:"gemma4",backend:"llamacpp",context_length:8192,file:"gemma-4-E4B-it-Q4_K_M.gguf",format:"gguf",hf_repo:"unsloth/gemma-4-E4B-it-GGUF",max_context_length:131072,parameters:7518069290,passthrough:true,quantization:"q4_k_m",revision:"bfc15c382204943c3a8fff0c750b94ae2364d7a3",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:8192,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Includes stored per-layer embeddings, not just the marketed effective parameter count. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/unsloth/gemma-4-E4B-it-GGUF/resolve/bfc15c382204943c3a8fff0c750b94ae2364d7a3/gemma-4-E4B-it-Q4_K_M.gguf","https://huggingface.co/google/gemma-4-E4B-it/blob/ee0ef6023621cff504d758262d4e04895a5af4a2/config.json","https://huggingface.co/google/gemma-4-E4B-it/blob/ee0ef6023621cff504d758262d4e04895a5af4a2/README.md"],weights_mb:4732}
48,sha256:"85a896a047553e842f25297ee5b031d64ff30147d9c4af17b1e4b394cd1fab87",size_bytes:4977171584}},voices:void 0,displayName:"Gemma 4 E4b",parameterLabel:"8B",sizeLabel:"4.6 GB",format:"gguf",quantization:"q4_k_m",hfRepo:"unsloth/gemma-4-E4B-it-GGUF",taskLabel:"Text Generation",familyLabel:"Google"},{id:"gemma3npc-1b",family:"google",task:"text-generation",parameters:1000000000,description:"Gemma3NPC 1B - Mobile-friendly NPC roleplay fine-tune of Gemma 3 1B for in-character game dialogue",defaultVariant:"universal",variants:{universal:{architecture:"gemma3",backend:"llamacpp",context_length:8192,file:"gemma-3npc-1b-q4_k_m.gguf",format:"gguf",hf_repo:"chimbiwide/Gemma3NPC-1b-Q4-GGUF",max_context_length:32768,parameters:999885952,passthrough:true,quantization:"q4_k_m",revision:"f50d2cb72e0180df56e3b29bd32e01de399e676a",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:8192,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/chimbiwide/Gemma3NPC-1b-Q4-GGUF/resolve/f50d2cb72e0180df56e3b29bd32e01de399e676a/gemma-3npc-1b-q4_k_m.gguf","https://huggingface.co/chimbiwide/Gemma3NPC-1b-float16/blob/c47f8b442fed7072806162de7837956f8dae320d/config.json","https://huggingface.co/chimbiwide/Gemma3NPC-1b-Q4-GGUF/blob/f50d2cb72e0180df56e3b29bd32e01de399e676a/README.md"],weights_mb:763},sha256:"f941dbe51e894579b0dcbb7bd4db16df10b707f062c4d496f992e21807386bc0",size_bytes:806058400}},voices:void 0,displayName:"Gemma3npc 1B",parameterLabel:"1B",sizeLabel:"768.7 MB",format:"gguf",quantization:"q4_k_m",hfRepo:"chimbiwide/Gemma3NPC-1b-Q4-GGUF",taskLabel:"Text Generation",familyLabel:"Google"},{id:"gemma3npc-it",family:"google",task:"text-generation",parameters:7000000000,description:"Gemma3NPC-it - NPC roleplay fine-tune of Gemma3n-E4B for in-character game dialogue",defaultVariant:"universal",variants:{universal:{architecture:"gemma3n",backend:"llamacpp",context_length:8192,file:"gemma-3npc-it-q4_k_m.gguf",format:"gguf",hf_repo:"chimbiwide/Gemma3NPC-it-Q4-GGUF",max_context_length:32768,parameters:6866172400,passthrough:true,quantization:"q4_k_m",revision:"ac156eee306a9437ac702a6354f7c884f0e4a949",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:8192,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Includes stored per-layer embeddings, not just the marketed effective parameter count. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/chimbiwide/Gemma3NPC-it-Q4-GGUF/resolve/ac156eee306a9437ac702a6354f7c884f0e4a949/gemma-3npc-it-q4_k_m.gguf","https://huggingface.co/chimbiwide/Gemma3NPC-it-float16/blob/a1125704f4b92be96fa5e2680769ae69efa09d91/config.json","https://huggingface.co/chimbiwide/Gemma3NPC-it-Q4-GGUF/blob/ac156eee306a9437ac702a6354f7c884f0e4a949/README.md","https://huggingface.co/chimbiwide/Gemma3NPC-it-Q8-GGUF/blob/834f8bab58ee0fe0d2a2191cd5f18c461b1d4a9f/README.md","https://huggingface.co/google/gemma-3n-E4B-it/blob/c1221e9c62e34a43ab7ffacd1be0ea71f126ef10/README.md","https://ai.google.dev/gemma/docs/gemma-3n"],weights_mb:4035},sha256:"8d1d1a7f693237fcc6515274d2a6665fe867471a1d95005a39e21ec1a3c8c384",size_bytes:4237062304}},voices:void 0,displayName:"Gemma3npc It",parameterLabel:"7B",sizeLabel:"3.9 GB",format:"gguf",quantization:"q4_k_m",hfRepo:"chimbiwide/Gemma3NPC-it-Q4-GGUF",taskLabel:"Text Generation",familyLabel:"Google"},{id:"kitten-tts-micro-0.8",family:"kittenml",task:"text-to-speech",parameters:40000000,description:"KittenTTS Micro 0.8 - Compact StyleTTS 2 (40M params, 8 named voices, OpenPhonemizer)",defaultVariant:"universal",variants:{universal:{architecture:"stylett
48s2",backend:"ort",file:"universal.xyb",format:"onnx",hf_repo:"xybrid-ai/KittenTTS-Micro-0.8-ONNX",parameters:40000000,quantization:"int8",revision:"82c994ebf73098eeea257f2003c313a3904ab249",runtime_requirements:{backend:"ort",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Includes the phonemizer model; expanded dictionary and voice assets are additional. Parameter count is publisher-rounded for the primary TTS model. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/xybrid-ai/KittenTTS-Micro-0.8-ONNX/resolve/82c994ebf73098eeea257f2003c313a3904ab249/universal.xyb","https://huggingface.co/api/models/xybrid-ai/KittenTTS-Micro-0.8-ONNX/revision/82c994ebf73098eeea257f2003c313a3904ab249?blobs=true","https://huggingface.co/api/models/KittenML/kitten-tts-micro-0.8/revision/1ccf72b2c2048fd17efac7de2fab32d10e225084?blobs=true","https://huggingface.co/KittenML/kitten-tts-micro-0.8/blob/1ccf72b2c2048fd17efac7de2fab32d10e225084/README.md"],weights_mb:99},sha256:"3e91e75d4f4daedeb2f7e9635e2a21468f4386fe8bf2fe1f24ebcd16fad2857d",size_bytes:94534715}},voices:{catalog:[{gender:"female",id:"expr-voice-2-f",language:"en",name:"Bella",style:"expressive"},{gender:"male",id:"expr-voice-2-m",language:"en",name:"Jasper",style:"expressive"},{gender:"female",id:"expr-voice-3-f",language:"en",name:"Luna",style:"expressive"},{gender:"male",id:"expr-voice-3-m",language:"en",name:"Bruno",style:"expressive"},{gender:"female",id:"expr-voice-4-f",language:"en",name:"Rosie",style:"expressive"},{gender:"male",id:"expr-voice-4-m",language:"en",name:"Hugo",style:"expressive"},{gender:"female",id:"expr-voice-5-f",language:"en",name:"Kiki",style:"expressive"},{gender:"male",id:"expr-voice-5-m",language:"en",name:"Leo",style:"expressive"}],default:"expr-voice-2-f"},displayName:"Kitten Tts Micro 0.8",parameterLabel:"40M",sizeLabel:"90.2 MB",format:"onnx",quantization:"int8",hfRepo:"xybrid-ai/KittenTTS-Micro-0.8-ONNX",taskLabel:"Text to Speech",familyLabel:"KittenML"},{id:"kitten-tts-mini-0.8",family:"kittenml",task:"text-to-speech",parameters:80000000,description:"KittenTTS Mini 0.8 - High-quality StyleTTS 2 (80M params, 8 named voices, OpenPhonemizer)",defaultVariant:"universal",variants:{universal:{architecture:"styletts2",backend:"ort",file:"universal.xyb",format:"onnx",hf_repo:"xybrid-ai/KittenTTS-Mini-0.8-ONNX",parameters:80000000,quantization:"int8",revision:"72ebd2af2ac4fcc762375264518a717c5e4cb57d",runtime_requirements:{backend:"ort",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Includes the phonemizer model; expanded dictionary and voice assets are additional. Parameter count is publisher-rounded for the primary TTS model. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/xybrid-ai/KittenTTS-Mini-0.8-ONNX/resolve/72ebd2af2ac4fcc762375264518a717c5e4cb57d/universal.xyb","https://huggingface.co/api/models/xybrid-ai/KittenTTS-Mini-0.8-ONNX/revision/72ebd2af2ac4fcc762375264518a717c5e4cb57d?blobs=true","https://huggingface.co/api/models/KittenML/kitten-tts-mini-0.8/revision/c02725660cea441db4c383af69f1f26f5cd00947?blobs=true","https://huggingface.co/KittenML/kitten-tts-mini-0.8/blob/c02725660cea441db4c383af69f1f26f5cd00947/README.md"],weights_mb:134},sha256:"801e3f17fca9c073339d36f681a70ec4cbda7ce010a3a2f2b98c5b9e1ff46987",size_bytes:116499044}},voices:{catalog:[{gender:"female",id:"expr-voice-2-f",language:"en",name:"Bella",style:"expressive"},{gender:"male",id:"expr-voice-2-m",language:"en",name:"Jasper",style:"expressive"},{gender:"female",id:"expr-voice-3-f",language:"en",name:"Luna",style:"expressive"},{gender:"male",id:"expr-voice-3-m",language:"en",name:"Bruno",style:"expressive"},{gender:"female",id:"expr-voice-4-f",language:"en",name:"Rosie",style:"expressive"},{gender:"male",id:"expr-voice-4-m",language:"en",name:"Hugo",style:"expressive"},{gender:"female",id:"expr-voice-5-f",language:"en",name:"Kiki",style:"expressive"},{gender:"male",id:"expr-voice-5-m",language:"en",name:"Leo",style:"expressive"}],default:"expr-voice-2-f"},displayName:"Kitten Tts Mini 0.8",parameterLabel:"80M",sizeLabel:"111.1 MB",format:"onnx",quantization:"int8",hfRepo:"xybrid-ai/KittenTTS-Mini-0.8-ONNX",taskLabel:"Text to Speech",fam
48ilyLabel:"KittenML"},{id:"kitten-tts-nano-0.2",family:"kittenml",task:"text-to-speech",parameters:15000000,description:"KittenTTS Nano 0.2 - Ultra-lightweight TTS (15M params, \u003C25MB)",defaultVariant:"universal",variants:{universal:{architecture:"kitten_tts",backend:"ort",file:"universal.xyb",format:"onnx",hf_repo:"xybrid-ai/KittenTTS-Nano-0.2-ONNX",parameters:15000000,quantization:"int8",revision:"5d773d00b4e424b355945099ecac539f975e93f0",runtime_requirements:{backend:"ort",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Parameter count is publisher-rounded for the primary TTS model. Dynamic int8 operators verified in the ONNX artifact; previous fp32 label was incorrect. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/xybrid-ai/KittenTTS-Nano-0.2-ONNX/resolve/5d773d00b4e424b355945099ecac539f975e93f0/universal.xyb","https://huggingface.co/api/models/xybrid-ai/KittenTTS-Nano-0.2-ONNX/revision/5d773d00b4e424b355945099ecac539f975e93f0?blobs=true","https://huggingface.co/api/models/KittenML/kitten-tts-nano-0.2/revision/9c81564aa56c6fb79f83780e87099357b88d6617?blobs=true","https://huggingface.co/KittenML/kitten-tts-nano-0.2/blob/9c81564aa56c6fb79f83780e87099357b88d6617/README.md"],weights_mb:23},sha256:"0a463fc55b772097a0fbfb77ed230763fd4de6c82dff897f6a4204eb5f099477",size_bytes:19239534}},voices:void 0,displayName:"Kitten Tts Nano 0.2",parameterLabel:"15M",sizeLabel:"18.3 MB",format:"onnx",quantization:"int8",hfRepo:"xybrid-ai/KittenTTS-Nano-0.2-ONNX",taskLabel:"Text to Speech",familyLabel:"KittenML"},{id:"kitten-tts-nano-0.8",family:"kittenml",task:"text-to-speech",parameters:15000000,description:"KittenTTS Nano 0.8 - Ultra-lightweight StyleTTS 2 (15M params, 8 named voices, OpenPhonemizer)",defaultVariant:"universal",variants:{universal:{architecture:"styletts2",backend:"ort",file:"universal.xyb",format:"onnx",hf_repo:"xybrid-ai/KittenTTS-Nano-0.8-ONNX",parameters:15000000,quantization:"int8",revision:"4a77bf67e01bb6137a0fbdf3c56eb057e33230b6",runtime_requirements:{backend:"ort",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Includes the phonemizer model; expanded dictionary and voice assets are additional. Parameter count is publisher-rounded for the primary TTS model. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/xybrid-ai/KittenTTS-Nano-0.8-ONNX/resolve/4a77bf67e01bb6137a0fbdf3c56eb057e33230b6/universal.xyb","https://huggingface.co/api/models/xybrid-ai/KittenTTS-Nano-0.8-ONNX/revision/4a77bf67e01bb6137a0fbdf3c56eb057e33230b6?blobs=true","https://huggingface.co/api/models/KittenML/kitten-tts-nano-0.8-fp32/revision/7a1db645b1f3ab9420761d87428e042b9cec3f26?blobs=true","https://huggingface.co/KittenML/kitten-tts-nano-0.8-fp32/blob/7a1db645b1f3ab9420761d87428e042b9cec3f26/README.md"],weights_mb:82},sha256:"da3ca1404842e0837662c5b40d358c6a5003101d70e03e8e876859660bd34ab4",size_bytes:82100441}},voices:void 0,displayName:"Kitten Tts Nano 0.8",parameterLabel:"15M",sizeLabel:"78.3 MB",format:"onnx",quantization:"int8",hfRepo:"xybrid-ai/KittenTTS-Nano-0.8-ONNX",taskLabel:"Text to Speech",familyLabel:"KittenML"},{id:"kokoro-82m",family:"hexgrad",task:"text-to-speech",parameters:82000000,description:"Kokoro 82M - High-quality TTS with 24 voices, Misaki dictionary",defaultVariant:"universal",variants:{universal:{architecture:"styletts2",backend:"ort",file:"universal.xyb",format:"onnx",hf_repo:"xybrid-ai/Kokoro-82M-v1.0-ONNX",parameters:82000000,quantization:"fp16",revision:"018ebc3be889c9b6bc6bd9fcb071cd893b7f6545",runtime_requirements:{backend:"ort",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/xybrid-ai/Kokoro-82M-v1.0-ONNX/resolve/018ebc3be889c9b6bc6bd9fcb
48071cd893b7f6545/universal.xyb","https://huggingface.co/api/models/xybrid-ai/Kokoro-82M-v1.0-ONNX/revision/018ebc3be889c9b6bc6bd9fcb071cd893b7f6545?blobs=true","https://huggingface.co/api/models/hexgrad/Kokoro-82M/revision/f3ff3571791e39611d31c381e3a41a3af07b4987?blobs=true","https://huggingface.co/hexgrad/Kokoro-82M/blob/f3ff3571791e39611d31c381e3a41a3af07b4987/README.md","https://huggingface.co/hexgrad/Kokoro-82M/blob/f3ff3571791e39611d31c381e3a41a3af07b4987/config.json"],weights_mb:170},sha256:"14bc36593ae65ab9c817f7796f5d6187cf750adcb767ec7478d6552d7440930e",size_bytes:183438685}},voices:{catalog:[{gender:"female",id:"af_heart",language:"en-US",name:"Heart",style:"neutral"},{gender:"female",id:"af_alloy",language:"en-US",name:"Alloy",style:"neutral"},{gender:"female",id:"af_aoede",language:"en-US",name:"Aoede",style:"neutral"},{gender:"female",id:"af_bella",language:"en-US",name:"Bella",style:"expressive"},{gender:"female",id:"af_jessica",language:"en-US",name:"Jessica",style:"neutral"},{gender:"female",id:"af_kore",language:"en-US",name:"Kore",style:"neutral"},{gender:"female",id:"af_nicole",language:"en-US",name:"Nicole",style:"whisper"},{gender:"female",id:"af_nova",language:"en-US",name:"Nova",style:"neutral"},{gender:"female",id:"af_river",language:"en-US",name:"River",style:"neutral"},{gender:"female",id:"af_sarah",language:"en-US",name:"Sarah",style:"neutral"},{gender:"female",id:"af_sky",language:"en-US",name:"Sky",style:"neutral"},{gender:"male",id:"am_adam",language:"en-US",name:"Adam",style:"neutral"},{gender:"male",id:"am_echo",language:"en-US",name:"Echo",style:"neutral"},{gender:"male",id:"am_eric",language:"en-US",name:"Eric",style:"neutral"},{gender:"male",id:"am_fenrir",language:"en-US",name:"Fenrir",style:"neutral"},{gender:"male",id:"am_liam",language:"en-US",name:"Liam",style:"neutral"},{gender:"male",id:"am_michael",language:"en-US",name:"Michael",style:"neutral"},{gender:"male",id:"am_onyx",language:"en-US",name:"Onyx",style:"neutral"},{gender:"male",id:"am_puck",language:"en-US",name:"Puck",style:"neutral"},{gender:"female",id:"bf_emma",language:"en-GB",name:"Emma",style:"neutral"},{gender:"female",id:"bf_isabella",language:"en-GB",name:"Isabella",style:"neutral"},{gender:"male",id:"bm_george",language:"en-GB",name:"George",style:"neutral"},{gender:"male",id:"bm_lewis",language:"en-GB",name:"Lewis",style:"neutral"},{gender:"male",id:"bm_daniel",language:"en-GB",name:"Daniel",style:"neutral"}],default:"af_heart"},displayName:"Kokoro 82M",parameterLabel:"82M",sizeLabel:"174.9 MB",format:"onnx",quantization:"fp16",hfRepo:"xybrid-ai/Kokoro-82M-v1.0-ONNX",taskLabel:"Text to Speech",familyLabel:"Hexgrad"},{id:"lfm2-vl-450m",family:"liquid",task:"vision-language",parameters:450000000,description:"Liquid LFM2-VL 450M compact vision-language model for edge inference.",defaultVariant:"universal",variants:{universal:{architecture:"lfm2",artifacts:[{file:"mmproj-LFM2-VL-450M-Q8_0.gguf",sha256:"a6a813fae4ba3f90852a990fdd4786a017dba0e80e599a366853ba539ee2b5ef",size_bytes:103890016}],backend:"llamacpp",context_length:4096,file:"LFM2-VL-450M-Q4_0.gguf",format:"gguf",hf_repo:"LiquidAI/LFM2-VL-450M-GGUF",max_context_length:32768,parameters:354483968,passthrough:true,quantization:"q4_0",revision:"7cf9679aa137039135e342aa77014f7b5f2bef73",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:4096,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Includes the registered vision projector; parameters describe the primary language model only. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/LiquidAI/LFM2-VL-450M-GGUF/resolve/7cf9679aa137039135e342aa77014f7b5f2bef73/LFM2-VL-450M-Q4_0.gguf","https://huggingface.co/LiquidAI/LFM2-VL-450M-GGUF/resolve/7cf9679aa137039135e342aa77014f7b5f2bef73/mmproj-LFM2-VL-450M-Q8_0.gguf","https://huggingface.co/LiquidAI/LFM2-VL-450M/tree/d355d88bf86e5c7054230232120e573ca2238286","https://huggingface.co/LiquidAI/LFM2-VL-450M/blob/d355d88bf86e5c7054230232120e573ca2238286/config.json","https://huggingface.co/LiquidAI/LFM2-VL-450M/blob/d355d88bf86e5c7054230232120e573ca2238286/README.md"],weights_mb:306},sha256:"1fbf4797e8669cdeabd97fa1d31ab046907cb1fc78b1c3d4bda1ba879574e5d7",size_bytes:219307424}},voices:void 0,displayName:"Lfm2 Vl 450M",parameterLabel:"450M",sizeLabel:"209.1 MB",format:"gguf",quantization:"q4_0",hfRepo:"LiquidAI/LFM2-VL-450M-GGUF",taskLabel:"vision-language",familyLabel:"Liquid AI"},{id:"lfm2.5-1.2b-instruct",family:"liquid",task:"text-generation",parameters:1170000000,description:"Liquid LFM2.5 1.2B Instru
48ct - hybrid conv+attention LLM optimized for edge deployment (agentic tasks, data extraction, RAG, tool calling)",defaultVariant:"universal",variants:{universal:{architecture:"lfm2",backend:"llamacpp",context_length:32768,file:"LFM2.5-1.2B-Instruct-Q4_K_M.gguf",format:"gguf",hf_repo:"LiquidAI/LFM2.5-1.2B-Instruct-GGUF",max_context_length:32768,parameters:1170340608,passthrough:true,quantization:"q4_k_m",revision:"6767265158422fb8a19c62ceb45f16f05363615b",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:32768,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct-GGUF/resolve/6767265158422fb8a19c62ceb45f16f05363615b/LFM2.5-1.2B-Instruct-Q4_K_M.gguf","https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct/tree/0f604ada3f766f9f257460c4c9f0b5d6f69d431b","https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct/blob/0f604ada3f766f9f257460c4c9f0b5d6f69d431b/config.json","https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct/blob/0f604ada3f766f9f257460c4c9f0b5d6f69d431b/README.md"],weights_mb:695},sha256:"b1b3de114215d9507409a662a501a631095a479a419584e8a2ded6304b19b4f5",size_bytes:730895168}},voices:void 0,displayName:"Lfm2.5 1.2B Instruct",parameterLabel:"1.2B",sizeLabel:"697.0 MB",format:"gguf",quantization:"q4_k_m",hfRepo:"LiquidAI/LFM2.5-1.2B-Instruct-GGUF",taskLabel:"Text Generation",familyLabel:"Liquid AI"},{id:"lfm2.5-1.2b-thinking",family:"liquid",task:"text-generation",parameters:1170000000,description:"Liquid LFM2.5 1.2B Thinking - hybrid conv+attention reasoning LLM optimized for edge deployment (fast CPU inference, low memory)",defaultVariant:"universal",variants:{universal:{architecture:"lfm2",backend:"llamacpp",context_length:32768,file:"LFM2.5-1.2B-Thinking-Q4_K_M.gguf",format:"gguf",hf_repo:"LiquidAI/LFM2.5-1.2B-Thinking-GGUF",max_context_length:32768,parameters:1170340608,passthrough:true,quantization:"q4_k_m",revision:"9584426effbfa44b677e925793a12188a36b44ce",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:32768,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/LiquidAI/LFM2.5-1.2B-Thinking-GGUF/resolve/9584426effbfa44b677e925793a12188a36b44ce/LFM2.5-1.2B-Thinking-Q4_K_M.gguf","https://huggingface.co/LiquidAI/LFM2.5-1.2B-Thinking/tree/f313478934a7612d22991f752959d7a1a8756fec","https://huggingface.co/LiquidAI/LFM2.5-1.2B-Thinking/blob/f313478934a7612d22991f752959d7a1a8756fec/config.json","https://huggingface.co/LiquidAI/LFM2.5-1.2B-Thinking/blob/f313478934a7612d22991f752959d7a1a8756fec/README.md"],weights_mb:695},sha256:"7223a2202405b02e8e1e6c5baa543c43dc98c1d9741a5c2a0ee1583212e1231b",size_bytes:730895360}},voices:void 0,displayName:"Lfm2.5 1.2B Thinking",parameterLabel:"1.2B",sizeLabel:"697.0 MB",format:"gguf",quantization:"q4_k_m",hfRepo:"LiquidAI/LFM2.5-1.2B-Thinking-GGUF",taskLabel:"Text Generation",familyLabel:"Liquid AI"},{id:"lfm2.5-230m",family:"liquid",task:"text-generation",parameters:230000000,description:"Liquid LFM2.5 230M - smallest hybrid conv+attention LLM optimized for edge deployment (9 languages, tool calling, fast CPU inference)",defaultVariant:"universal",variants:{universal:{architecture:"lfm2",backend:"llamacpp",context_length:32768,file:"LFM2.5-230M-Q4_K_M.gguf",format:"gguf",hf_repo:"LiquidAI/LFM2.5-230M-GGUF",max_context_length:32768,parameters:229693184,passthrough:true,quantization:"q4_k_m",revision:"cdf97bd8205908758f44aec508d68ac1aef98f5c",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:32768,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB;
48 actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/LiquidAI/LFM2.5-230M-GGUF/resolve/cdf97bd8205908758f44aec508d68ac1aef98f5c/LFM2.5-230M-Q4_K_M.gguf","https://huggingface.co/LiquidAI/LFM2.5-230M/tree/40cb2ad3b3044d5a41eee083a6103c8b523afa45","https://huggingface.co/LiquidAI/LFM2.5-230M/blob/40cb2ad3b3044d5a41eee083a6103c8b523afa45/config.json","https://huggingface.co/LiquidAI/LFM2.5-230M/blob/40cb2ad3b3044d5a41eee083a6103c8b523afa45/README.md"],weights_mb:145},sha256:"7bbd90384d3deffe4c646ec9643b212802d32d4ce417c90a1ec9282100650062",size_bytes:153406304}},voices:void 0,displayName:"Lfm2.5 230M",parameterLabel:"230M",sizeLabel:"146.3 MB",format:"gguf",quantization:"q4_k_m",hfRepo:"LiquidAI/LFM2.5-230M-GGUF",taskLabel:"Text Generation",familyLabel:"Liquid AI"},{id:"lfm2.5-350m",family:"liquid",task:"text-generation",parameters:354000000,description:"Liquid LFM2.5 350M - hybrid conv+attention LLM optimized for edge deployment (9 languages, tool calling)",defaultVariant:"universal",variants:{universal:{architecture:"lfm2",backend:"llamacpp",context_length:32768,file:"LFM2.5-350M-Q4_K_M.gguf",format:"gguf",hf_repo:"LiquidAI/LFM2.5-350M-GGUF",max_context_length:32768,parameters:354483968,passthrough:true,quantization:"q4_k_m",revision:"9969000761ce34de907bf20017cbfc3d52d6eaf9",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:32768,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/LiquidAI/LFM2.5-350M-GGUF/resolve/9969000761ce34de907bf20017cbfc3d52d6eaf9/LFM2.5-350M-Q4_K_M.gguf","https://huggingface.co/LiquidAI/LFM2.5-350M/tree/9e6c6ccf47cd318696e137d381a7ded8fe4df09f","https://huggingface.co/LiquidAI/LFM2.5-350M/blob/9e6c6ccf47cd318696e137d381a7ded8fe4df09f/config.json","https://huggingface.co/LiquidAI/LFM2.5-350M/blob/9e6c6ccf47cd318696e137d381a7ded8fe4df09f/README.md"],weights_mb:217},sha256:"7e6f72643caafc9a68256686638c4d7916f2cec76d1df478d4c3ddcd95a6aed4",size_bytes:229312224}},voices:void 0,displayName:"Lfm2.5 350M",parameterLabel:"354M",sizeLabel:"218.7 MB",format:"gguf",quantization:"q4_k_m",hfRepo:"LiquidAI/LFM2.5-350M-GGUF",taskLabel:"Text Generation",familyLabel:"Liquid AI"},{id:"lfm2.5-vl-3b",family:"liquid",task:"vision-language",parameters:3000000000,description:"Liquid LFM2.5-VL 3B vision-language model for local inference.",defaultVariant:"universal",variants:{universal:{architecture:"lfm2",artifacts:[{file:"mmproj-LFM2.5-VL-3B-Q8_0.gguf",sha256:"8ba27050dc88737db66b856d3b74e0e6cf54bee35fa4d9d9808f69ee556bbd43",size_bytes:583109120}],backend:"llamacpp",context_length:4096,file:"LFM2.5-VL-3B-Q4_K_M.gguf",format:"gguf",hf_repo:"LiquidAI/LFM2.5-VL-3B-GGUF",max_context_length:32768,parameters:2697198592,passthrough:true,quantization:"q4_k_m",revision:"3e0e828198e2abb75a957ad823f5d691c13f0f28",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:4096,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Includes the registered vision projector; parameters describe the primary language model only. Pinned to the revision matching both published hashes; current publisher advertises 32768 despite the older GGUF header saying 128000. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/LiquidAI/LFM2.5-VL-3B-GGUF/resolve/3e0e828198e2abb75a957ad823f5d691c13f0f28/LFM2.5-VL-3B-Q4_K_M.gguf","https://huggingface.co/LiquidAI/LFM2.5-VL-3B-GGUF/resolve/3e0e828198e2abb75a957ad823f5d691c13f0f28/mmproj-LFM2.5-VL-3B-Q8_0.gguf","https://huggingface.co/LiquidAI/LFM2.5-VL-3B/tree/a3af5799199acdd2a4f56ac4342816abb46c12a9","https://huggingface.co/LiquidAI/LFM2.5-VL-3B/blob/a3af5799199acdd2a4f56ac4342816abb46c12a9/config.json","https://huggingface.co/LiquidAI/LFM2.5-VL-3B/blob/a3af5799199acdd2a4f56ac4342816abb46c12a9/README.md","https://huggingface.co/api/models/LiquidAI/LFM2.5-VL-3B-GGUF/revision/3e
480e828198e2abb75a957ad823f5d691c13f0f28?blobs=true"],weights_mb:2146},sha256:"83c18dfba02c75769cdd63f73e37c343400e82d434ff1b14bcc1cb02fcf2f5f2",size_bytes:1674454240}},voices:void 0,displayName:"Lfm2.5 Vl 3B",parameterLabel:"3B",sizeLabel:"1.6 GB",format:"gguf",quantization:"q4_k_m",hfRepo:"LiquidAI/LFM2.5-VL-3B-GGUF",taskLabel:"vision-language",familyLabel:"Liquid AI"},{id:"llama-3.2-1b",family:"meta",task:"text-generation",parameters:1000000000,description:"Llama 3.2 1B - Meta's lightweight mobile-optimized instruction-tuned LLM with 128K context",defaultVariant:"universal",variants:{universal:{architecture:"llama",backend:"llamacpp",context_length:8192,file:"Llama-3.2-1B-Instruct-Q4_K_M.gguf",format:"gguf",hf_repo:"bartowski/Llama-3.2-1B-Instruct-GGUF",max_context_length:131072,parameters:1235814400,passthrough:true,quantization:"q4_k_m",revision:"067b946cf014b7c697f3654f621d577a3e3afd1c",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:8192,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/bartowski/Llama-3.2-1B-Instruct-GGUF/resolve/067b946cf014b7c697f3654f621d577a3e3afd1c/Llama-3.2-1B-Instruct-Q4_K_M.gguf","https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct/blob/9213176726f574b556790deb65791e0c5aa438b6/config.json","https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct/blob/9213176726f574b556790deb65791e0c5aa438b6/README.md","https://huggingface.co/api/models/meta-llama/Llama-3.2-1B-Instruct/revision/9213176726f574b556790deb65791e0c5aa438b6"],weights_mb:763},sha256:"6f85a640a97cf2bf5b8e764087b1e83da0fdb51d7c9fab7d0fece9385611df83",size_bytes:807694464}},voices:void 0,displayName:"Llama 3.2 1B",parameterLabel:"1B",sizeLabel:"770.3 MB",format:"gguf",quantization:"q4_k_m",hfRepo:"bartowski/Llama-3.2-1B-Instruct-GGUF",taskLabel:"Text Generation",familyLabel:"Meta"},{id:"ministral-3-3b",family:"mistral",task:"text-generation",parameters:3400000000,description:"Ministral 3 3B Instruct - Mistral AI's edge-optimized instruction-tuned LLM with 256K context",defaultVariant:"universal",variants:{universal:{architecture:"mistral3",backend:"llamacpp",context_length:4096,file:"Ministral-3-3B-Instruct-2512-Q4_K_M.gguf",format:"gguf",hf_repo:"mistralai/Ministral-3-3B-Instruct-2512-GGUF",max_context_length:262144,parameters:3429006336,passthrough:true,quantization:"q4_k_m",revision:"eb599d408350ea2bb60452cb86be7c7b2fc28227",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:4096,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512-GGUF/resolve/eb599d408350ea2bb60452cb86be7c7b2fc28227/Ministral-3-3B-Instruct-2512-Q4_K_M.gguf","https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512/blob/b35d4dfe56c142746f54dbd64f579faab2744308/config.json","https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512/blob/b35d4dfe56c142746f54dbd64f579faab2744308/README.md","https://huggingface.co/api/models/mistralai/Ministral-3-3B-Instruct-2512/revision/b35d4dfe56c142746f54dbd64f579faab2744308"],weights_mb:2040},sha256:"9ed150d4367e68df0ac8e1540f6ddc65b42d0ee26378329d1ecbca60f93fc5f8",size_bytes:2147023008}},voices:void 0,displayName:"Ministral 3 3B",parameterLabel:"3.4B",sizeLabel:"2.0 GB",format:"gguf",quantization:"q4_k_m",hfRepo:"mistralai/Ministral-3-3B-Instruct-2512-GGUF",taskLabel:"Text Generation",familyLabel:"Mistral"},{id:"mistral-7b",family:"mistral",task:"text-generation",parameters:7000000000,description:"Mistral 7B Instruct v0.3 - High-quality desktop LLM with function calling support",defaultVariant:"universal",variants:{universal:{architecture:"llama",backend:"llamacpp",context_length:8192,file:"Mistral-7B-Instruct-v0.3-Q4_K_M.gguf",format:"gguf",hf_repo:"bartowski/Mistral-7B-Instruct-v0.3-GGUF",max_context_length:32768,parameters:7248023552,passthrough:true,quantization:"q4_k_m",revision:"61fd416
487fff3ab01ee1cfe0da183fa27a944db48",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:8192,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/bartowski/Mistral-7B-Instruct-v0.3-GGUF/resolve/61fd4167fff3ab01ee1cfe0da183fa27a944db48/Mistral-7B-Instruct-v0.3-Q4_K_M.gguf","https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3/blob/c170c708c41dac9275d15a8fff4eca08d52bab71/config.json","https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3/blob/c170c708c41dac9275d15a8fff4eca08d52bab71/README.md","https://huggingface.co/api/models/mistralai/Mistral-7B-Instruct-v0.3/revision/c170c708c41dac9275d15a8fff4eca08d52bab71"],weights_mb:4170},sha256:"1270d22c0fbb3d092fb725d4d96c457b7b687a5f5a715abe1e818da303e562b6",size_bytes:4372812000}},voices:void 0,displayName:"Mistral 7B",parameterLabel:"7B",sizeLabel:"4.1 GB",format:"gguf",quantization:"q4_k_m",hfRepo:"bartowski/Mistral-7B-Instruct-v0.3-GGUF",taskLabel:"Text Generation",familyLabel:"Mistral"},{id:"neutts-air-q4",family:"neuphonic",task:"text-to-speech",parameters:500000000,description:"NeuTTS Air (~500M) — codec TTS with voice cloning. Q4 GGUF backbone + NeuCodec decoder. Higher quality than Nano.",defaultVariant:"universal",variants:{universal:{architecture:"qwen2",backend:"llamacpp",context_length:2048,file:"universal.xyb",format:"gguf",hf_repo:"xybrid-ai/NeuTTS-Air-Q4",max_context_length:32768,parameters:552914304,quantization:"q4_0",revision:"2d53a4294fd992f3a55a291f8725d19f34e8d573",runtime_requirements:{backend:"llamacpp",confidence:"low",context_length:2048,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Includes ORT codec and phonemizer serialized model bytes; dictionary, voice/reference assets and audio buffers need additional memory. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/xybrid-ai/NeuTTS-Air-Q4/resolve/2d53a4294fd992f3a55a291f8725d19f34e8d573/universal.xyb","https://huggingface.co/api/models/xybrid-ai/NeuTTS-Air-Q4/revision/2d53a4294fd992f3a55a291f8725d19f34e8d573?blobs=true","https://huggingface.co/api/models/neuphonic/neutts-air/revision/3bf9403646f0ca46a3f2250355d2ce4d1f70559c?blobs=true","https://huggingface.co/neuphonic/neutts-air/blob/3bf9403646f0ca46a3f2250355d2ce4d1f70559c/README.md","https://huggingface.co/neuphonic/neutts-air/blob/3b58b776406b62fdc137e31ea53d728f5c22a4ed/config.json","https://huggingface.co/xybrid-ai/NeuTTS-Air-Q4/resolve/2d53a4294fd992f3a55a291f8725d19f34e8d573/neutts-air-Q4_0.gguf"],weights_mb:852},sha256:"4916949ad1143d78a309a35a03a8d01b4dc607cd27f2bcc4f45ddc830f9212e4",size_bytes:791642564}},voices:{catalog:[{gender:"female",id:"jo",language:"en-US",name:"Jo",style:"neutral"},{gender:"male",id:"dave",language:"en-US",name:"Dave",style:"neutral"}],default:"jo"},displayName:"Neutts Air Q4",parameterLabel:"500M",sizeLabel:"755.0 MB",format:"gguf",quantization:"q4_0",hfRepo:"xybrid-ai/NeuTTS-Air-Q4",taskLabel:"Text to Speech",familyLabel:"Neuphonic"},{id:"neutts-nano-q4",family:"neuphonic",task:"text-to-speech",parameters:120000000,description:"NeuTTS Nano (120M) — codec TTS with voice cloning. Q4 GGUF backbone + NeuCodec ONNX decoder.",defaultVariant:"universal",variants:{universal:{architecture:"llama",backend:"llamacpp",context_length:2048,file:"universal.xyb",format:"gguf",hf_repo:"xybrid-ai/NeuTTS-Nano-Q4",max_context_length:2048,parameters:228704832,quantization:"q4_0",revision:"a941568eabe1bc9df9c1151ae67964eef08e37ee",runtime_requirements:{backend:"llamacpp",confidence:"low",context_length:2048,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB;
48 actual residency depends on backend/offload. Includes ORT codec and phonemizer serialized model bytes; dictionary, voice/reference assets and audio buffers need additional memory. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/xybrid-ai/NeuTTS-Nano-Q4/resolve/a941568eabe1bc9df9c1151ae67964eef08e37ee/universal.xyb","https://huggingface.co/api/models/xybrid-ai/NeuTTS-Nano-Q4/revision/a941568eabe1bc9df9c1151ae67964eef08e37ee?blobs=true","https://huggingface.co/api/models/neuphonic/neutts-nano/revision/ca2ba957804f0d80dd38ee819bdbd07311f95d15?blobs=true","https://huggingface.co/neuphonic/neutts-nano/blob/ca2ba957804f0d80dd38ee819bdbd07311f95d15/README.md","https://huggingface.co/neuphonic/neutts-nano/blob/411c78b8eb75a1871ea4e6b7a755fa2642afe1d5/config.json","https://huggingface.co/xybrid-ai/NeuTTS-Nano-Q4/resolve/a941568eabe1bc9df9c1151ae67964eef08e37ee/neutts-nano-Q4_0.gguf"],weights_mb:533},sha256:"29df32108c97ebf45edaab9b5d61356f91dea03220114fce61f4ce52c2193dc9",size_bytes:492557989}},voices:{catalog:[{gender:"female",id:"jo",language:"en",name:"Jo",style:"neutral"},{gender:"male",id:"dave",language:"en",name:"Dave",style:"neutral"},{gender:"female",id:"emily",language:"en",name:"Emily",style:"neutral"},{gender:"male",id:"paul",language:"en",name:"Paul",style:"neutral"},{gender:"female",id:"sophie",language:"en",name:"Sophie",style:"neutral"},{gender:"male",id:"steven",language:"en",name:"Steven",style:"neutral"}],default:"jo"},displayName:"Neutts Nano Q4",parameterLabel:"120M",sizeLabel:"469.7 MB",format:"gguf",quantization:"q4_0",hfRepo:"xybrid-ai/NeuTTS-Nano-Q4",taskLabel:"Text to Speech",familyLabel:"Neuphonic"},{id:"phi4-mini",family:"microsoft",task:"text-generation",parameters:3800000000,description:"Phi-4 Mini 3.8B - Microsoft's compact reasoning LLM with 128K context",defaultVariant:"universal",variants:{universal:{architecture:"phi3",backend:"llamacpp",context_length:8192,file:"microsoft_Phi-4-mini-instruct-Q4_K_M.gguf",format:"gguf",hf_repo:"bartowski/microsoft_Phi-4-mini-instruct-GGUF",max_context_length:131072,parameters:3836021760,passthrough:true,quantization:"q4_k_m",revision:"7ff82c2aaa4dde30121698a973765f39be5288c0",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:8192,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF/resolve/7ff82c2aaa4dde30121698a973765f39be5288c0/microsoft_Phi-4-mini-instruct-Q4_K_M.gguf","https://huggingface.co/microsoft/Phi-4-mini-instruct/blob/cfbefacb99257ffa30c83adab238a50856ac3083/config.json","https://huggingface.co/microsoft/Phi-4-mini-instruct/blob/cfbefacb99257ffa30c83adab238a50856ac3083/README.md","https://huggingface.co/api/models/microsoft/Phi-4-mini-instruct/revision/cfbefacb99257ffa30c83adab238a50856ac3083"],weights_mb:2369},sha256:"01999f17c39cc3074afae5e9c539bc82d45f2dd7faa3917c66cbef76fce8c0c2",size_bytes:2491874688}},voices:void 0,displayName:"Phi4 Mini",parameterLabel:"3.8B",sizeLabel:"2.3 GB",format:"gguf",quantization:"q4_k_m",hfRepo:"bartowski/microsoft_Phi-4-mini-instruct-GGUF",taskLabel:"Text Generation",familyLabel:"Microsoft"},{id:"qwen2.5-0.5b-instruct",family:"qwen",task:"text-generation",parameters:500000000,description:"Qwen 2.5 0.5B Instruct - Small but capable instruction-tuned LLM from Alibaba Cloud",defaultVariant:"universal",variants:{universal:{architecture:"qwen2",backend:"llamacpp",context_length:4096,file:"qwen2.5-0.5b-instruct-q4_k_m.gguf",format:"gguf",hf_repo:"Qwen/Qwen2.5-0.5B-Instruct-GGUF",max_context_length:32768,parameters:494032768,passthrough:true,quantization:"q4_k_m",revision:"9217f5db79a29953eb74d5343926648285ec7e67",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:4096,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB;
48 actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct-GGUF/resolve/9217f5db79a29953eb74d5343926648285ec7e67/qwen2.5-0.5b-instruct-q4_k_m.gguf","https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct/blob/7ae557604adf67be50417f59c2c2f167def9a775/config.json","https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct/blob/7ae557604adf67be50417f59c2c2f167def9a775/README.md","https://huggingface.co/api/models/Qwen/Qwen2.5-0.5B-Instruct/revision/7ae557604adf67be50417f59c2c2f167def9a775"],weights_mb:463},sha256:"74a4da8c9fdbcd15bd1f6d01d621410d31c6fc00986f5eb687824e7b93d7a9db",size_bytes:491400032}},voices:void 0,displayName:"Qwen2.5 0.5B Instruct",parameterLabel:"500M",sizeLabel:"468.6 MB",format:"gguf",quantization:"q4_k_m",hfRepo:"Qwen/Qwen2.5-0.5B-Instruct-GGUF",taskLabel:"Text Generation",familyLabel:"Qwen"},{id:"qwen3.5-0.8b",family:"qwen",task:"text-generation",parameters:800000000,description:"Qwen 3.5 0.8B - Alibaba Cloud's lightweight multimodal LLM (text-only mode, 201 languages)",defaultVariant:"universal",variants:{universal:{architecture:"qwen35",backend:"llamacpp",context_length:4096,file:"universal.xyb",format:"gguf",hf_repo:"xybrid-ai/Qwen3.5-0.8B-GGUF",max_context_length:262144,parameters:752393024,quantization:"q4_k_m",revision:"9fd84d9b913ec0d45b460877c41afaee0ea739ea",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:4096,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Text-only bundle. Runtime context 4096 is the audited SDK serde default; embedded loose metadata 262144 is the model maximum. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/xybrid-ai/Qwen3.5-0.8B-GGUF/resolve/9fd84d9b913ec0d45b460877c41afaee0ea739ea/universal.xyb","https://huggingface.co/unsloth/Qwen3.5-0.8B-GGUF/resolve/6ab461498e2023f6e3c1baea90a8f0fe38ab64d0/Qwen3.5-0.8B-Q4_K_M.gguf","https://huggingface.co/Qwen/Qwen3.5-0.8B/tree/2fc06364715b967f1860aea9cf38778875588b17","https://huggingface.co/Qwen/Qwen3.5-0.8B/blob/2fc06364715b967f1860aea9cf38778875588b17/config.json","https://huggingface.co/Qwen/Qwen3.5-0.8B/blob/2fc06364715b967f1860aea9cf38778875588b17/README.md","https://github.com/xybrid-ai/xybrid/blob/0d610601d300a5519124523eccdc26aee09d2dd1/crates/xybrid-core/src/execution/template/metadata.rs#L585"],weights_mb:498},sha256:"f4fcf5d5fcf6aad7d24d99b8c6b1b4b575b225b621229a47c9fb4c7e97c1a16f",size_bytes:519807056}},voices:void 0,displayName:"Qwen3.5 0.8B",parameterLabel:"800M",sizeLabel:"495.7 MB",format:"gguf",quantization:"q4_k_m",hfRepo:"xybrid-ai/Qwen3.5-0.8B-GGUF",taskLabel:"Text Generation",familyLabel:"Qwen"},{id:"qwen3.5-2b",family:"qwen",task:"text-generation",parameters:2000000000,description:"Qwen 3.5 2B - Alibaba Cloud's compact multimodal LLM (text-only mode, 201 languages)",defaultVariant:"universal",variants:{universal:{architecture:"qwen35",backend:"llamacpp",context_length:4096,file:"universal.xyb",format:"gguf",hf_repo:"xybrid-ai/Qwen3.5-2B-GGUF",max_context_length:262144,parameters:1881825088,quantization:"q4_k_m",revision:"c86ae8d451aa5f526888ad084baa86cd6c480550",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:4096,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Text-only bundle. Runtime context 4096 is the audited SDK serde default; embedded loose metadata 262144 is the model maximum. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/xybrid-ai/Qwen3.5-2B-GGUF/blob/c86ae8d451aa5f526888ad084baa86cd6c480550/universal.xyb","https://huggingface.co/unsloth/Qwen3.5-2B-GGUF/resolve/f6d5376be1edb4d4
4816d56da11e5397a961aca8ae/Qwen3.5-2B-Q4_K_M.gguf","https://huggingface.co/Qwen/Qwen3.5-2B/tree/15852e8c16360a2fea060d615a32b45270f8a8fc","https://huggingface.co/Qwen/Qwen3.5-2B/blob/15852e8c16360a2fea060d615a32b45270f8a8fc/config.json","https://huggingface.co/Qwen/Qwen3.5-2B/blob/15852e8c16360a2fea060d615a32b45270f8a8fc/README.md","https://github.com/xybrid-ai/xybrid/blob/0d610601d300a5519124523eccdc26aee09d2dd1/crates/xybrid-core/src/execution/template/metadata.rs#L585"],weights_mb:1212},sha256:"ef0d23ab630e53eed1413d3ed45f07184a188efd7bcc449be8d45f274bf561ad",size_bytes:1259689856}},voices:void 0,displayName:"Qwen3.5 2B",parameterLabel:"2B",sizeLabel:"1.2 GB",format:"gguf",quantization:"q4_k_m",hfRepo:"xybrid-ai/Qwen3.5-2B-GGUF",taskLabel:"Text Generation",familyLabel:"Qwen"},{id:"smollm2-360m",family:"huggingface",task:"text-generation",parameters:360000000,description:"SmolLM2 360M - HuggingFace's best tiny LLM, excellent quality/size ratio",defaultVariant:"universal",variants:{universal:{architecture:"llama",backend:"llamacpp",context_length:4096,file:"SmolLM2-360M-Instruct-Q4_K_M.gguf",format:"gguf",hf_repo:"bartowski/SmolLM2-360M-Instruct-GGUF",max_context_length:8192,parameters:361821120,passthrough:true,quantization:"q4_k_m",revision:"7be6f65f1db715fe5dc5a4634c0d459b4eed42ec",runtime_requirements:{backend:"llamacpp",confidence:"medium",context_length:4096,notes:"Weight baseline is GGUF tensor storage, rounded up to MiB; actual residency depends on backend/offload. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/bartowski/SmolLM2-360M-Instruct-GGUF/resolve/7be6f65f1db715fe5dc5a4634c0d459b4eed42ec/SmolLM2-360M-Instruct-Q4_K_M.gguf","https://huggingface.co/HuggingFaceTB/SmolLM2-360M-Instruct/blob/a10cc1512eabd3dde888204e902eca88bddb4951/config.json","https://huggingface.co/HuggingFaceTB/SmolLM2-360M-Instruct/blob/a10cc1512eabd3dde888204e902eca88bddb4951/README.md","https://huggingface.co/api/models/HuggingFaceTB/SmolLM2-360M-Instruct/revision/a10cc1512eabd3dde888204e902eca88bddb4951"],weights_mb:257},sha256:"2fa3f013dcdd7b99f9b237717fa0b12d75bbb89984cc1274be1471a465bac9c2",size_bytes:270590880},web:{architecture:"llama",backend:"litertlm",context_length:4096,file:"SmolLM2_360M_instruct.litertlm",format:"litertlm",hf_repo:"litert-community/SmolLM2-360M-Instruct",max_context_length:8192,model_metadata:{description:"SmolLM2 360M - HuggingFace's best tiny LLM, excellent quality/size ratio",execution_template:{context_length:4096,model_file:"SmolLM2_360M_instruct.litertlm",type:"LiteRtLm"},files:["SmolLM2_360M_instruct.litertlm"],metadata:{architecture:"llama",instruction_tuned:true,languages:["en"],notes:"Best tiny 
48LLM available. Uses llama.cpp backend.",task:"text-generation"},model_id:"smollm2-360m",postprocessing:[],preprocessing:[],version:"1.0"},parameters:361821120,passthrough:true,quantization:"int8",revision:"507c99cfe6541ba2bcd84818786f7b025935e5e1",sha256:"8e2834da211b439751af968ed650febdde5a8cb8d88bc6c1a3059f049caa5c2e",size_bytes:373719040}},voices:void 0,displayName:"Smollm2 360M",parameterLabel:"360M",sizeLabel:"258.1 MB",format:"gguf",quantization:"q4_k_m",hfRepo:"bartowski/SmolLM2-360M-Instruct-GGUF",taskLabel:"Text Generation",familyLabel:"HuggingFace"},{id:"wav2vec2-base-960h",family:"facebook",task:"speech-recognition",parameters:95000000,description:"Wav2Vec2 Base 960h - English ASR with CTC decoding",defaultVariant:"universal",variants:{universal:{architecture:"wav2vec2",backend:"ort",file:"universal.xyb",format:"onnx",hf_repo:"xybrid-ai/wav2vec2-base-960h",parameters:94395552,quantization:"fp32",revision:"e5ef7280456bd9579460e928cb7254d485944ef1",runtime_requirements:{backend:"ort",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/xybrid-ai/wav2vec2-base-960h/resolve/e5ef7280456bd9579460e928cb7254d485944ef1/universal.xyb","https://huggingface.co/api/models/xybrid-ai/wav2vec2-base-960h/revision/e5ef7280456bd9579460e928cb7254d485944ef1?blobs=true","https://huggingface.co/api/models/facebook/wav2vec2-base-960h/revision/22aad52d435eb6dbaf354bdad9b0da84ce7d6156?blobs=true","https://huggingface.co/facebook/wav2vec2-base-960h/blob/22aad52d435eb6dbaf354bdad9b0da84ce7d6156/README.md","https://huggingface.co/api/models/onnx-community/wav2vec2-base-960h-ONNX/revision/729c1a6730fb549c20a1c73a3d3f96f11020225e?blobs=true","https://huggingface.co/onnx-community/wav2vec2-base-960h-ONNX/blob/729c1a6730fb549c20a1c73a3d3f96f11020225e/config.json"],weights_mb:361},sha256:"cfe868703d23c61eafbc80603d53dc80a45eb4684dbd23189420b817ce7cd68a",size_bytes:231028388}},voices:void 0,displayName:"Wav2vec2 Base 960H",parameterLabel:"95M",sizeLabel:"220.3 MB",format:"onnx",quantization:"fp32",hfRepo:"xybrid-ai/wav2vec2-base-960h",taskLabel:"Speech Recognition",familyLabel:"Meta"},{id:"whisper-base-en-ggml",family:"openai",task:"speech-recognition",parameters:74000000,description:"Whisper Base English - English-only ASR (whisper.cpp/GGML runtime)",defaultVariant:"universal",variants:{universal:{architecture:"whisper",backend:"whispercpp",file:"ggml-base.en-q5_1.bin",format:"ggml",hf_repo:"ggerganov/whisper.cpp",parameters:72593408,passthrough:true,quantization:"q5_1",revision:"5359861c739e955e79d9a303bcbc70fb988958b1",runtime_requirements:{backend:"whispercpp",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Audio encoder positions and decoder limits are not LLM context lengths; input duration affects additional memory. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/ggerganov/whisper.cpp/resolve/5359861c739e955e79d9a303bcbc70fb988958b1/ggml-base.en-q5_1.bin","https://huggingface.co/api/models/ggerganov/whisper.cpp/revision/5359861c739e955e79d9a303bcbc70fb988958b1?blobs=true","https://huggingface.co/api/models/openai/whisper-base.en/revision/911407f4214e0e1d82085af863093ec0b66f9cd6?blobs=true","https://huggingface.co/openai/whisper-base.en/blob/911407f4214e0e1d82085af863093ec0b66f9cd6/README.md","https://huggingface.co/openai/whisper-base.en/blob/911407f4214e0e1d82085af863093ec0b66f9cd6/config.json"],weights_mb:57},sha256:"4baf70dd0d7c4247ba2b81fafd9c01005ac77c2f9ef064e00dcf195d0e2fdd2f",size_bytes:59721011}},voices:void 0,displayName:"Whisper Base En Ggml",parameterLabel:"74M",sizeLabel:"57.0 MB",format:"ggml",quantization:"q5_1",hfRepo:"ggerganov/whisper.cpp",taskLabel:"Speech Recognition",familyLabel:"OpenAI"},{id:"whisper-base-ggml",family:"openai",task:"speech-recognition",parameters:74000000,description:"Whisper Base - Multilingual ASR (whisper.cpp/GGML runtime)",defaultVariant:"universal",variants:{universal:{architecture:"whisper",backend:"whispercpp",file:"ggml-base-q5_1.bin",format:"ggml",hf_repo:"ggerganov/whisper.cpp",parameters:72593920,passthrough:true,quantization:"q5_1",revision:"5359861c739e955e79d9a303bcbc70fb988958b1",runtime_requirements:{backend:"whispercpp",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Audio encoder positions and decoder limits are not LLM context lengths; input duration affects additional memory. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/ggerganov/whisper.cpp/resolve/5359861c739e955e79d9a303bcbc70fb988958b1/ggml-base-q5_1.bin","https://huggingface.co/api/models/ggerganov/whisper.cpp/revision/5359861c739e955e79d9a303bcbc70fb988958b1?blobs=true","https://huggingface.co/api/models/openai/whisper-base/revision/e37978b90ca9030d5170a5c07aadb050351a65bb?blobs=true","https://huggingface.co/openai/whisper-base/blob/e37978b90ca9030d5170a5c07aadb050351a65bb/README.md","https://huggingface.co/openai/whisper-base/blob/e37978b90ca9030d5170a5c07aadb050351a65bb/config.json"],weights_mb:57},sha256:"422f1ae452ade6f30a004d7e5c6a43195e4433bc370bf23fac9cc591f01a8898",size_bytes:59707625}},voices:void 0,displayName:"Whisper Base Ggml",parameterLabel:"74M",sizeLabel:"56.9 MB",format:"ggml",quantization:"q5_1",hfRepo:"ggerganov/whisper.cpp",taskLabel:"Speech Recognition",familyLabel:"OpenAI"},{id:"whisper-large-v3-turbo-ggml",family:"openai",task:"speech-recognition",parameters:809000000,description:"Whisper Large v3 Turbo - Highest-accuracy multilingual ASR (whisper.cpp/GGML runtime)",defaultVariant:"universal",variants:{universal:{architecture:"whisper",backend:"whispercpp",file:"ggml-large-v3-turbo-q5_0.bin",format:"ggml",hf_repo:"ggerganov/whisper.cpp",parameters:808878080,passthrough:true,quantization:"q5_0",revision:"5359861c739e955e79d9a303bcbc70fb988958b1",runtime_requirements:{backend:"whispercpp",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Audio encoder positions and decoder limits are not LLM context lengths; input duration affects additional memory. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/ggerganov/whisper.cpp/resolve/5359861c739e955e79d9a303bcbc70fb988958b1/ggml-large-v3-turbo-q5_0.bin","https://huggingface.co/api/models/ggerganov/whisper.cpp/revision/5359861c739e955e79d9a303bcbc70fb988958b1?blobs=true","https://huggingface.co/api/models/openai/whisper-large-v3-turbo/revision/41f01f3fe87f28c78e2fbf8b568835947dd65ed9?blobs=true","https://huggingface.co/openai/whisper-large-v3-turbo/blob/41f01f3fe87f28c78e2fbf8b568835947dd65ed9/README.md","https://huggingface.co/openai/whisper-large-v3-turbo/blob/41f01f3fe87f28c78e2fbf8b568835947dd65ed9/config.json"],weights_mb:548},sha256:"394221709cd5ad1f40c46e6031ca61bce88931e6e088c188294c6d5a55ffa7e2",size_bytes:574041195}},voices:void 0,displayName:"Whisper Large V3 Turbo Ggml",parameterLabel:"809M",sizeLabel:"547.4 MB",format:"ggml",quantization:"q5_0",hfRepo:"ggerganov/whisper.cpp",taskLabel:"Speech Recognition",familyLabel:"OpenAI"},{id:"whisper-medium-ggml",family:"openai",task:"speech-recognition",parameters:769000000,description:"Whisper Medium - Multilingual ASR (whisper.cpp/GGML runtime)",defaultVariant:"universal",variants:{universal:{architecture:"whisper",backend:"whispercpp",file:"ggml-medium-q5_0.bin",format:"ggml",hf_repo:"ggerganov/whisper.cpp",parameters:763857920,passthrough:true,quantization:"q5_0",revision:"5359861c739e955e79d9a303bcbc70fb988958b1",runtime_requirements:{backend:"whispercpp",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Audio encoder positions and decoder limits are not LLM context lengths; input duration affects additional memory. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/ggerganov/whisper.cpp/resolve/5359861c739e955e79d9a303bcbc70fb988958b1/ggml-medium-q5_0.bin","https://huggingface.co/api/models/ggerganov/whisper.cpp/revision/5359861c739e955e79d9a303bcbc70fb988958b1?blobs=true","https://huggingface.co/api/models/openai/whisper-medium/revision/abdf7c39ab9d0397620ccaea8974cc764cd0953e?blobs=true","https://huggingface.co/openai/whisper-medium/blob/abdf7c39ab9d0397620ccaea8974cc764cd0953e/README.md","https://huggingface.co/openai/whisper-medium/blob/abdf7c39ab9d0397620ccaea8974cc764cd0953e/config.json"],weights_mb:515},sha256:"19fea4b380c3a618ec4723c3eef2eb785ffba0d0538cf43f8f235e7b3b34220f",size_bytes:539212467}},voices:void 0,displayName:"Whisper Medium Ggml",parameterLabel:"769M",sizeLabel:"514.2 MB",format:"ggml",quantization:"q5_0",hfRepo:"ggerganov/whisper.cpp",taskLabel:"Speech Recognition",familyLabel:"OpenAI"},{id:"whisper-small-ggml",family:"openai",task:"speech-recognition",parameters:244000000,description:"Whisper Small - Multilingual ASR (whisper.cpp/GGML runtime)",defaultVariant:"universal",variants:{universal:{architecture:"whisper",backend:"whispercpp",file:"ggml-small-q5_1.bin",format:"ggml",hf_repo:"ggerganov/whisper.cpp",parameters:241734912,passthrough:true,quantization:"q5_1",revision:"5359861c739e955e79d9a303bcbc70fb988958b1",runtime_requirements:{backend:"whispercpp",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Audio encoder positions and decoder limits are not LLM context lengths; input duration affects additional memory. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/ggerganov/whisper.cpp/resolve/5359861c739e955e79d9a303bcbc70fb988958b1/ggml-small-q5_1.bin","https://huggingface.co/api/models/ggerganov/whisper.cpp/revision/5359861c739e955e79d9a303bcbc70fb988958b1?blobs=true","https://huggingface.co/api/models/openai/whisper-small/revision/973afd24965f72e36ca33b3055d56a652f456b4d?blobs=true","https://huggingface.co/openai/whisper-small/blob/973afd24965f72e36ca33b3055d56a652f456b4d/README.md","https://huggingface.co/openai/whisper-small/blob/973afd24965f72e36ca33b3055d56a652f456b4d/config.json"],weights_mb:182},sha256:"ae85e4a935d7a567bd102fe55afc16bb595bdb618e11b2fc7591bc08120411bb",size_bytes:190085487}},voices:void 0,displayName:"Whisper Small Ggml",parameterLabel:"244M",sizeLabel:"181.3 MB",format:"ggml",quantization:"q5_1",hfRepo:"ggerganov/whisper.cpp",taskLabel:"Speech Recognition",familyLabel:"OpenAI"},{id:"whisper-tiny",family:"openai",task:"speech-recognition",parameters:39000000,description:"Whisper Tiny - Fast multilingual ASR (Candle/SafeTensors runtime)",defaultVariant:"universal",variants:{universal:{architecture:"whisper",backend:"candle",file:"universal.xyb",format:"safetensors",hf_repo:"xybrid-ai/whisper-tiny",parameters:37760640,quantization:"fp32",revision:"a1458152adaa12d26f797153753b096b64721769",runtime_requirements:{backend:"candle",confidence:"low",notes:"Weight baseline is the SafeTensors F32 payload; Candle Whisper loads F32. Audio/mel buffers, caches and activations are additional. Audio encoder positions and decoder limits are not LLM context lengths; input duration affects additional memory. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/xybrid-ai/whisper-tiny/resolve/a1458152adaa12d26f797153753b096b64721769/universal.xyb","https://huggingface.co/api/models/xybrid-ai/whisper-tiny/revision/a1458152adaa12d26f797153753b096b64721769?blobs=true","https://huggingface.co/api/models/openai/whisper-tiny/revision/169d4a4341b33bc18d8881c4b69c2e104e1cc0af?blobs=true","https://huggingface.co/openai/whisper-tiny/blob/169d4a4341b33bc18d8881c4b69c2e104e1cc0af/README.md","https://huggingface.co/openai/whisper-tiny/blob/169d4a4341b33bc18d8881c4b69c2e104e1cc0af/config.json"],weights_mb:145},sha256:"8c8f569ce51130bb1eea697581d376f538c9e8239e08ed43d93078beb1a55c99",size_bytes:88842231}},voices:void 0,displayName:"Whisper Tiny",parameterLabel:"39M",sizeLabel:"84.7 MB",format:"safetensors",quantization:"fp32",hfRepo:"xybrid-ai/whisper-tiny",taskLabel:"Speech Recognition",familyLabel:"OpenAI"},{id:"whisper-tiny-en-ggml",family:"openai",task:"speech-recognition",parameters:39000000,description:"Whisper Tiny English - Fastest English-only ASR (whisper.cpp/GGML runtime)",defaultVariant:"universal",variants:{universal:{architecture:"whisper",backend:"whispercpp",file:"ggml-tiny.en-q5_1.bin",format:"ggml",hf_repo:"ggerganov/whisper.cpp",parameters:37760256,passthrough:true,quantization:"q5_1",revision:"5359861c739e955e79d9a303bcbc70fb988958b1",runtime_requirements:{backend:"whispercpp",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Audio encoder positions and decoder limits are not LLM context lengths; input duration affects additional memory. Total runtime RAM is unknown: cache, compute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/ggerganov/whisper.cpp/resolve/5359861c739e955e79d9a303bcbc70fb988958b1/ggml-tiny.en-q5_1.bin","https://huggingface.co/api/models/ggerganov/whisper.cpp/revision/5359861c739e955e79d9a303bcbc70fb988958b1?blobs=true","https://huggingface.co/api/models/openai/whisper-tiny.en/revision/87c7102498dcde7456f24cfd30239ca606ed9063?blobs=true","https://huggingface.co/openai/whisper-tiny.en/blob/87c7102498dcde7456f24cfd30239ca606ed9063/README.md","https://huggingface.co/openai/whisper-tiny.en/blob/87c7102498dcde7456f24cfd30239ca606ed9063/config.json"],weights_mb:31},sha256:"c77c5766f1cef09b6b7d47f21b546cbddd4157886b3b5d6d4f709e91e66c7c2b",size_bytes:32166155}},voices:void 0,displayName:"Whisper Tiny En Ggml",parameterLabel:"39M",sizeLabel:"30.7 MB",format:"ggml",quantization:"q5_1",hfRepo:"ggerganov/whisper.cpp",taskLabel:"Speech Recognition",familyLabel:"OpenAI"},{id:"whisper-tiny-ggml",family:"openai",task:"speech-recognition",parameters:39000000,description:"Whisper Tiny - Fast multilingual ASR (whisper.cpp/GGML runtime)",defaultVariant:"universal",variants:{universal:{architecture:"whisper",backend:"whispercpp",file:"ggml-tiny-q5_1.bin",format:"ggml",hf_repo:"ggerganov/whisper.cpp",parameters:37760640,passthrough:true,quantization:"q5_1",revision:"5359861c739e955e79d9a303bcbc70fb988958b1",runtime_requirements:{backend:"whispercpp",confidence:"low",notes:"Approximate weight baseline from uncompressed serialized model components, including graph/header bytes; not an ONNX/GGML allocation measurement. Audio encoder positions and decoder limits are not LLM context lengths; input duration affects additional memory. Total runtime RAM is unknown: cache, com
48pute workspace, activations and application/device overhead are not bounded. Missing minimum/recommended memory must yield Unknown compatibility.",source:"estimated",source_urls:["https://huggingface.co/ggerganov/whisper.cpp/resolve/5359861c739e955e79d9a303bcbc70fb988958b1/ggml-tiny-q5_1.bin","https://huggingface.co/api/models/ggerganov/whisper.cpp/revision/5359861c739e955e79d9a303bcbc70fb988958b1?blobs=true","https://huggingface.co/api/models/openai/whisper-tiny/revision/169d4a4341b33bc18d8881c4b69c2e104e1cc0af?blobs=true","https://huggingface.co/openai/whisper-tiny/blob/169d4a4341b33bc18d8881c4b69c2e104e1cc0af/README.md","https://huggingface.co/openai/whisper-tiny/blob/169d4a4341b33bc18d8881c4b69c2e104e1cc0af/config.json"],weights_mb:31},sha256:"818710568da3ca15689e31a743197b520007872ff9576237bda97bd1b469c3d7",size_bytes:32152673}},voices:void 0,displayName:"Whisper Tiny Ggml",parameterLabel:"39M",sizeLabel:"30.7 MB",format:"ggml",quantization:"q5_1",hfRepo:"ggerganov/whisper.cpp",taskLabel:"Speech Recognition",familyLabel:"OpenAI"}],seo:{title:"Model Registry | Xybrid",description:"Browse 37 on-device AI models for text generation, text-to-speech, and speech recognition. Run locally on iOS, Android, macOS, and more.",url:"https://www.xybrid.ai/models"}},uses:{}}],
49							form: null,
50							error: null
51						});
52					});
53				}
54			</script>
54
55		</div>
56	</body>
57</html>

Line numbers count LF bytes from the start of the resource, as the search results do. Vendor segments are library code the classifier recognised; they are stored but not indexed. Bytes are shown as Latin1 characters, one per byte.