{
  "slug": "glm-5-3-flash",
  "order": 99,
  "name": "GLM-5.3-Flash (320B)",
  "baseTitle": "GLM-5.3-Flash",
  "company": "Z.ai",
  "country": "China",
  "flag": "🇨🇳",
  "date": "2026-08-26",
  "decoderType": "Sparse hybrid",
  "attention": "3:1 Kimi Delta Attention and MLA with DeepSeek Sparse Attention",
  "layerMix": "34 KDA + 11 MLA/DSA + 1 MTP layer",
  "scale": "320B total, 18B active (5.6% active)",
  "totalParams": 320000000000,
  "activeParams": 18000000000,
  "context": "1,048,576",
  "contextTokens": 1048576,
  "license": "MIT License",
  "kvCacheText": "11 KiB",
  "kvCacheKiB": 11,
  "kvCacheLevel": "Very low",
  "summary": "Natively multimodal GLM refresh, previously previewed as Ox Alpha, that replaces most MLA/DSA layers with Kimi Delta Attention and adds four-stream mHC.",
  "keyDetail": "Mixes 34 Kimi Delta Attention layers with 11 MLA/DSA layers and routes residuals through four mHC streams.",
  "image": {
    "full": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/glm-5.3-flash.webp",
    "thumb": "https://sebastianraschka.com/llm-architecture-gallery/images/architectures/thumbnails/glm-5.3-flash.webp",
    "alt": "Figure 1: GLM-5.3-Flash",
    "width": 960,
    "height": 804
  },
  "articleUrl": null,
  "links": [
    {
      "label": "config.json",
      "href": "https://huggingface.co/zai-org/GLM-5.3-Flash/blob/main/config.json",
      "title": "zai-org/GLM-5.3-Flash"
    },
    {
      "label": "License",
      "href": "https://huggingface.co/zai-org/GLM-5.3-Flash/blob/main/LICENSE",
      "title": null
    },
    {
      "label": "Tech report",
      "href": "https://arxiv.org/pdf/2602.15763",
      "title": null
    }
  ],
  "concepts": [
    {
      "slug": "hybrid-attention",
      "label": "Hybrid Attention",
      "name": "Hybrid Attention",
      "url": "https://llm-atlas-7or.pages.dev/concepts/hybrid-attention/"
    },
    {
      "slug": "shortconv",
      "label": "ShortConv",
      "name": "Short Convolution",
      "url": "https://llm-atlas-7or.pages.dev/concepts/shortconv/"
    },
    {
      "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": "nope",
      "label": "NoPE",
      "name": "No Positional Embeddings",
      "url": "https://llm-atlas-7or.pages.dev/concepts/nope/"
    },
    {
      "slug": "mhc",
      "label": "mHC",
      "name": "Manifold-constrained Hyper-Connections",
      "url": "https://llm-atlas-7or.pages.dev/concepts/mhc/"
    },
    {
      "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": {
    "v5": {
      "label": "AA Intelligence Index v5.0",
      "total": 27.8,
      "profile": {
        "Knowledge Work": 43.7,
        "Coding": 23.7,
        "Science": 12.5,
        "General": 30.9
      },
      "url": "https://artificialanalysis.ai/models/glm-5-3-flash"
    },
    "v4": {
      "label": "AA Intelligence Index v4.3.2",
      "total": 41.8,
      "profile": {
        "General": 39.7,
        "Scientific": 27.6,
        "Coding": 42.2,
        "Agents": 53
      },
      "url": "https://artificialanalysis.ai/models/glm-5-3-flash"
    }
  },
  "sourceUrl": "https://sebastianraschka.com/llm-architecture-gallery/#card-glm-5-3-flash",
  "family": "hybrid-moe",
  "url": "https://llm-atlas-7or.pages.dev/models/glm-5-3-flash/",
  "markdownUrl": "https://llm-atlas-7or.pages.dev/models/glm-5-3-flash.md"
}