{
  "slug": "tencent-hy4-preview-770b-a49b",
  "order": 101,
  "name": "Tencent Hy4-preview (770B-A49B)",
  "baseTitle": "Tencent Hy4-preview 770B-A49B",
  "company": "Tencent",
  "country": "China",
  "flag": "🇨🇳",
  "date": "2026-08-28",
  "decoderType": "Sparse MoE",
  "attention": "Gated MLA with DeepSeek Sparse Attention and IndexCache",
  "layerMix": "78 gated MLA/DSA + 1 MTP layer",
  "scale": "770B backbone, 49B active (6.4% active); plus 10B MTP",
  "totalParams": 770000000000,
  "activeParams": 49000000000,
  "context": "1,048,576",
  "contextTokens": 1048576,
  "license": "Apache License 2.0",
  "kvCacheText": "87.8 KiB",
  "kvCacheKiB": 87.8,
  "kvCacheLevel": "Moderate",
  "summary": "Tencent's 770B sparse MoE combines gated DeepSeek Sparse Attention, cross-layer index reuse, and four-stream identity Hyper-Connections.",
  "keyDetail": "Uses 256 routed experts with 8 routed plus 1 shared expert active per token, a top-2,048 DSA indexer, cross-layer index reuse, and four iHC residual streams.",
  "image": {
    "full": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/tencent-hy4.webp",
    "thumb": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/thumbnails/tencent-hy4.webp",
    "alt": "Figure 1: Tencent Hy4-preview 770B-A49B",
    "width": 960,
    "height": 739
  },
  "articleUrl": null,
  "links": [
    {
      "label": "config.json",
      "href": "https://huggingface.co/tencent/Hy4-preview/blob/main/config.json",
      "title": "tencent/Hy4-preview"
    },
    {
      "label": "License",
      "href": "https://huggingface.co/tencent/Hy4-preview/blob/main/LICENSE",
      "title": null
    }
  ],
  "concepts": [
    {
      "slug": "mhc",
      "label": "mHC",
      "name": "Manifold-constrained Hyper-Connections",
      "url": "https://llm-atlas-7or.pages.dev/concepts/mhc/"
    },
    {
      "slug": "mla",
      "label": "MLA",
      "name": "Multi-Head Latent Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/mla/"
    },
    {
      "slug": "deepseek-sparse-attention",
      "label": "DSA",
      "name": "DeepSeek Sparse Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/deepseek-sparse-attention/"
    },
    {
      "slug": "indexshare",
      "label": "IndexShare",
      "name": "IndexShare",
      "url": "https://llm-atlas-7or.pages.dev/concepts/indexshare/"
    },
    {
      "slug": "gated-attention",
      "label": "Gated Attention",
      "name": "Gated Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/gated-attention/"
    },
    {
      "slug": "moe",
      "label": "MoE",
      "name": "Mixture of Experts",
      "url": "https://llm-atlas-7or.pages.dev/concepts/moe/"
    },
    {
      "slug": "mtp",
      "label": "MTP",
      "name": "Multi-Token Prediction",
      "url": "https://llm-atlas-7or.pages.dev/concepts/mtp/"
    }
  ],
  "aai": null,
  "sourceUrl": "https://sebastianraschka.com/llm-architecture-gallery/#card-tencent-hy4-preview-770b-a49b",
  "family": "moe",
  "url": "https://llm-atlas-7or.pages.dev/models/tencent-hy4-preview-770b-a49b/",
  "markdownUrl": "https://llm-atlas-7or.pages.dev/models/tencent-hy4-preview-770b-a49b.md"
}