{
  "slug": "laguna-s-2-1",
  "order": 89,
  "name": "Laguna S 2.1 (118B)",
  "baseTitle": "Laguna S 2.1",
  "company": "Poolside",
  "country": "United States",
  "flag": "🇺🇸",
  "date": "2026-07-21",
  "decoderType": "Sparse MoE",
  "attention": "Gated GQA with QK-Norm and 3:1 sliding-window/global attention",
  "layerMix": "36 sliding-window + 12 global GQA",
  "scale": "118B total, 8B active (6.8% active)",
  "totalParams": 118000000000,
  "activeParams": 8000000000,
  "context": "1,048,576",
  "contextTokens": 1048576,
  "license": "OpenMDW License 1.1",
  "kvCacheText": "192 KiB",
  "kvCacheKiB": 192,
  "kvCacheLevel": "High",
  "summary": "Poolside's 118B sparse MoE scales the Laguna XS recipe to 48 layers and 1M context while activating about 8B parameters per token.",
  "keyDetail": "Uses 256 routed experts with top-10 routing plus one shared expert, a dense first layer, and per-layer query-head counts.",
  "image": {
    "full": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/laguna-s-2-1.webp",
    "thumb": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/thumbnails/laguna-s-2-1.webp",
    "alt": "Figure 1: Laguna S 2.1",
    "width": 960,
    "height": 913
  },
  "articleUrl": null,
  "links": [
    {
      "label": "config.json",
      "href": "https://huggingface.co/poolside/Laguna-S-2.1/blob/main/config.json",
      "title": "poolside/Laguna-S-2.1"
    },
    {
      "label": "License",
      "href": "https://huggingface.co/poolside/Laguna-S-2.1/blob/main/LICENSE.md",
      "title": null
    },
    {
      "label": "Tech report",
      "href": "https://poolside.ai/blog/introducing-laguna-s-2-1",
      "title": null
    }
  ],
  "concepts": [
    {
      "slug": "gqa",
      "label": "GQA",
      "name": "Grouped-Query Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/gqa/"
    },
    {
      "slug": "qk-norm",
      "label": "QK-Norm",
      "name": "QK-Norm",
      "url": "https://llm-atlas-7or.pages.dev/concepts/qk-norm/"
    },
    {
      "slug": "swa",
      "label": "SWA",
      "name": "Sliding Window Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/swa/"
    },
    {
      "slug": "moe",
      "label": "MoE",
      "name": "Mixture of Experts",
      "url": "https://llm-atlas-7or.pages.dev/concepts/moe/"
    },
    {
      "slug": "attention-budgeting",
      "label": "Attention Budgeting",
      "name": "Layer-wise attention budgeting",
      "url": "https://llm-atlas-7or.pages.dev/concepts/attention-budgeting/"
    }
  ],
  "aai": null,
  "sourceUrl": "https://sebastianraschka.com/llm-architecture-gallery/#card-laguna-s-2-1",
  "family": "moe",
  "url": "https://llm-atlas-7or.pages.dev/models/laguna-s-2-1/",
  "markdownUrl": "https://llm-atlas-7or.pages.dev/models/laguna-s-2-1.md"
}