{
  "slug": "qwen3-8-flash-next-125b-a6b",
  "order": 100,
  "name": "Qwen3.8-Flash-Next (125B-A6B)",
  "baseTitle": "Qwen3.8-Flash-Next 125B-A6B",
  "company": "Alibaba Cloud",
  "country": "China",
  "flag": "🇨🇳",
  "date": "2026-08-26",
  "decoderType": "Sparse hybrid",
  "attention": "3:1 Gated DeltaNet and Qwen Sparse Attention (QSA)",
  "layerMix": "12 QSA + 36 DeltaNet",
  "scale": "125B core, 6B active (4.8% active); plus 51B n-gram embeddings and 4B MTP",
  "totalParams": 125000000000,
  "activeParams": 6000000000,
  "context": "262,144",
  "contextTokens": 262144,
  "license": "Qwen Community License 1.0",
  "kvCacheText": "24 KiB",
  "kvCacheKiB": 24,
  "kvCacheLevel": "Very low",
  "summary": "Experimental vision-language MoE whose text backbone combines Gated DeltaNet, micro-block QSA, four-branch gated residuals, and n-gram embeddings; the figure omits the vision encoder and MTP head.",
  "keyDetail": "Replaces full attention with micro-block QSA while using four gated residual branches and a 20M-entry n-gram embedding at layer 2.",
  "image": {
    "full": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/qwen3-8-flash-next.webp",
    "thumb": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/thumbnails/qwen3-8-flash-next.webp",
    "alt": "Figure 1: Qwen3.8-Flash-Next 125B-A6B",
    "width": 960,
    "height": 798
  },
  "articleUrl": null,
  "links": [
    {
      "label": "config.json",
      "href": "https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/main/config.json",
      "title": "Qwen/Qwen3.8-Flash-Next"
    },
    {
      "label": "License",
      "href": "https://huggingface.co/Qwen/Qwen3.8-Flash-Next/blob/main/LICENSE",
      "title": null
    },
    {
      "label": "Tech report",
      "href": "https://github.com/QwenLM/Qwen3.8-Flash-Next/blob/main/tech_report.pdf",
      "title": null
    }
  ],
  "concepts": [
    {
      "slug": "hybrid-attention",
      "label": "Hybrid Attention",
      "name": "Hybrid Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/hybrid-attention/"
    },
    {
      "slug": "qwen-sparse-attention",
      "label": "Qwen Sparse Attention",
      "name": "Qwen Sparse Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/qwen-sparse-attention/"
    },
    {
      "slug": "gated-residuals",
      "label": "Gated Residuals",
      "name": "Gated Residuals",
      "url": "https://llm-atlas-7or.pages.dev/concepts/gated-residuals/"
    },
    {
      "slug": "gqa",
      "label": "GQA",
      "name": "Grouped-Query Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/gqa/"
    },
    {
      "slug": "gated-attention",
      "label": "Gated Attention",
      "name": "Gated Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/gated-attention/"
    },
    {
      "slug": "per-layer-embeddings",
      "label": "PLE",
      "name": "Per-Layer Embeddings",
      "url": "https://llm-atlas-7or.pages.dev/concepts/per-layer-embeddings/"
    },
    {
      "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": "mtp",
      "label": "MTP",
      "name": "Multi-Token Prediction",
      "url": "https://llm-atlas-7or.pages.dev/concepts/mtp/"
    }
  ],
  "aai": {
    "v5": {
      "label": "AA Intelligence Index v5.0",
      "total": null,
      "profile": {
        "General": 26.6
      },
      "url": null
    },
    "v4": {
      "label": "AA Intelligence Index v4.3.2",
      "total": 39.8,
      "profile": {
        "General": 35.8,
        "Scientific": 24.6,
        "Coding": 37.9,
        "Agents": 55.1
      },
      "url": "https://artificialanalysis.ai/models/qwen3-8-flash-next"
    }
  },
  "sourceUrl": "https://sebastianraschka.com/llm-architecture-gallery/#card-qwen3-8-flash-next-125b-a6b",
  "family": "hybrid-moe",
  "url": "https://llm-atlas-7or.pages.dev/models/qwen3-8-flash-next-125b-a6b/",
  "markdownUrl": "https://llm-atlas-7or.pages.dev/models/qwen3-8-flash-next-125b-a6b.md"
}