{
  "slug": "longcat-flash-lite-68-5b-a3b",
  "order": 54,
  "name": "LongCat-Flash-Lite (68.5B-A3B)",
  "baseTitle": "LongCat-Flash-Lite 68.5B-A3B",
  "company": "Meituan",
  "country": "China",
  "flag": "🇨🇳",
  "date": "2026-01-28",
  "decoderType": "Sparse MoE",
  "attention": "MLA with RoPE + NoPE",
  "layerMix": "14 logical layers / 28 MLA sublayers",
  "scale": "68.5B total, ~3B active avg (4.4% active)",
  "totalParams": 68500000000,
  "activeParams": 3000000000,
  "context": "256,000",
  "contextTokens": 256000,
  "license": "MIT License",
  "kvCacheText": "31.5 KiB",
  "kvCacheKiB": 31.5,
  "kvCacheLevel": "Low",
  "summary": "Meituan's LongCat-Next text backbone is a lightweight MoE that shifts a large share of parameters into N-gram embeddings while keeping sparse inference around 3B active parameters.",
  "keyDetail": "Uses N-gram embedding expansion, Shortcut MoE routing over 256 routed plus 128 zero/identity experts, and top-k 12 expert choices per token.",
  "image": {
    "full": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/longcat-flash-lite-68-5b-a3b.webp",
    "thumb": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/thumbnails/longcat-flash-lite-68-5b-a3b.webp",
    "alt": "Figure 1: LongCat-Flash-Lite 68.5B-A3B",
    "width": 879,
    "height": 960
  },
  "articleUrl": null,
  "links": [
    {
      "label": "config.json",
      "href": "https://huggingface.co/meituan-longcat/LongCat-Flash-Lite/blob/main/config.json",
      "title": "meituan-longcat/LongCat-Flash-Lite"
    },
    {
      "label": "Tech report",
      "href": "https://arxiv.org/abs/2601.21204",
      "title": null
    }
  ],
  "concepts": [
    {
      "slug": "mla",
      "label": "MLA",
      "name": "Multi-Head Latent Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/mla/"
    },
    {
      "slug": "moe",
      "label": "MoE",
      "name": "Mixture of Experts",
      "url": "https://llm-atlas-7or.pages.dev/concepts/moe/"
    }
  ],
  "aai": {
    "v4": {
      "label": "AA Intelligence Index v4.3.2",
      "total": 11.5,
      "profile": {
        "Scientific": 2.9
      },
      "url": "https://artificialanalysis.ai/models/longcat-flash-lite"
    }
  },
  "sourceUrl": "https://sebastianraschka.com/llm-architecture-gallery/#card-longcat-flash-lite-68-5b-a3b",
  "family": "moe",
  "url": "https://llm-atlas-7or.pages.dev/models/longcat-flash-lite-68-5b-a3b/",
  "markdownUrl": "https://llm-atlas-7or.pages.dev/models/longcat-flash-lite-68-5b-a3b.md"
}