{
  "slug": "kolibri-1-78b",
  "order": 108,
  "name": "Kolibri-1 (78B)",
  "baseTitle": "Kolibri-1 78B",
  "company": "Aleph Alpha",
  "country": "Germany",
  "flag": "🇩🇪",
  "date": "2026-10-03",
  "decoderType": "Sparse MoE",
  "attention": "GQA with QK-Norm and 4:1 sliding-window/global attention",
  "layerMix": "40 sliding-window + 10 global GQA",
  "scale": "78B total, 3.46B active (4.4% active)",
  "totalParams": 78000000000,
  "activeParams": 3460000000,
  "context": "262,144 native; 1,048,576 extrapolated",
  "contextTokens": 262144,
  "license": "Apache License 2.0",
  "kvCacheText": "100 KiB",
  "kvCacheKiB": 100,
  "kvCacheLevel": "Moderate",
  "summary": "Aleph Alpha's German-English reasoning model uses 50 MoE blocks with a 4:1 mix of sliding-window and global GQA.",
  "keyDetail": "Uses top-6 routing over 384 experts plus one shared expert, with sandwich RMSNorm, RoPE in local layers, and NoPE in global layers.",
  "image": {
    "full": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/kolibri-1.webp",
    "thumb": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/thumbnails/kolibri-1.webp",
    "alt": "Figure 1: Kolibri-1 78B",
    "width": 960,
    "height": 951
  },
  "articleUrl": null,
  "links": [
    {
      "label": "config.json",
      "href": "https://huggingface.co/Aleph-Alpha/Kolibri-1/blob/main/config.json",
      "title": "Aleph-Alpha/Kolibri-1"
    },
    {
      "label": "Tech report",
      "href": "https://aleph-alpha.com/downloads/tech-report.pdf",
      "title": null
    }
  ],
  "concepts": [
    {
      "slug": "gqa",
      "label": "GQA",
      "name": "Grouped-Query Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/gqa/"
    },
    {
      "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": "qk-norm",
      "label": "QK-Norm",
      "name": "QK-Norm",
      "url": "https://llm-atlas-7or.pages.dev/concepts/qk-norm/"
    },
    {
      "slug": "nope",
      "label": "NoPE",
      "name": "No Positional Embeddings",
      "url": "https://llm-atlas-7or.pages.dev/concepts/nope/"
    }
  ],
  "aai": null,
  "sourceUrl": "https://sebastianraschka.com/llm-architecture-gallery/#card-kolibri-1-78b",
  "family": "moe",
  "url": "https://llm-atlas-7or.pages.dev/models/kolibri-1-78b/",
  "markdownUrl": "https://llm-atlas-7or.pages.dev/models/kolibri-1-78b.md"
}