{
 "id": "glm-4.5-air",
 "name": "GLM-4.5-Air",
 "repo": "zai-org/GLM-4.5-Air",
 "lab": {
  "name": "Z.ai",
  "country": "China"
 },
 "released": "2025-07-20",
 "parameters": 110.468824832,
 "activeParameters": 13,
 "maxContext": 131072,
 "architecture": "46 layers · 46 full attention (8 KV × 128)",
 "page": "https://studiotvai.com/vram-requirements/glm-4-5-air",
 "calculator": "https://studiotvai.com/llm-vram-calculator?model=glm-4.5-air",
 "weightsGiB": {
  "bf16": 199.03,
  "fp8": 100.67,
  "q4_k_m": 64.69
 },
 "kvCacheGiBPerRequest": [
  {
   "context": 8192,
   "bf16": 1.438,
   "fp8": 0.719
  },
  {
   "context": 32768,
   "bf16": 5.75,
   "fp8": 2.875
  },
  {
   "context": 131072,
   "bf16": 23,
   "fp8": 11.5
  }
 ],
 "recurrentStateGiBPerRequest": 0,
 "gpusNeeded": {
  "scenario": "one request of 8192 tokens, BF16 KV cache; vLLM on data-center GPUs, llama.cpp elsewhere",
  "gpus": [
   {
    "gpu": "rtx-4090",
    "name": "GeForce RTX 4090",
    "memoryGB": 24,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 16,
     "fp8": 5,
     "q4_k_m": 4
    }
   },
   {
    "gpu": "rtx-5090",
    "name": "GeForce RTX 5090",
    "memoryGB": 32,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 8,
     "fp8": 4,
     "q4_k_m": 3
    }
   },
   {
    "gpu": "rtx-pro-6000",
    "name": "RTX PRO 6000 Blackwell",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 3,
     "fp8": 2,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "mac-m5-max-128",
    "name": "Mac M5 Max 128 GB",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 3,
     "fp8": 2,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "h100",
    "name": "NVIDIA H100",
    "memoryGB": 80,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 3,
     "fp8": 2,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "h200",
    "name": "NVIDIA H200",
    "memoryGB": 141,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 2,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "b200",
    "name": "NVIDIA B200 (HGX)",
    "memoryGB": 180,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 2,
     "fp8": 1,
     "q4_k_m": 1
    }
   }
  ]
 },
 "ggufFiles": {
  "repo": "unsloth/GLM-4.5-Air-GGUF",
  "gib": {
   "BF16": 205.82,
   "Q2_K": 42.21,
   "Q2_K_L": 42.34,
   "UD-IQ1_M": 37.31,
   "UD-IQ1_S": 35.86,
   "UD-IQ2_M": 41.34,
   "UD-IQ2_XXS": 39.75,
   "UD-Q2_K_XL": 44.19,
   "UD-TQ1_0": 35.63,
   "IQ4_NL": 58.4,
   "IQ4_XS": 56.33,
   "Q3_K_M": 53.28,
   "Q3_K_S": 49.04,
   "Q4_0": 58.47,
   "Q4_1": 64.64,
   "Q4_K_M": 67.96,
   "Q4_K_S": 62.43,
   "Q5_K_M": 77.78,
   "Q5_K_S": 73.03,
   "Q6_K": 92.21,
   "Q8_0": 109.39,
   "UD-IQ3_XXS": 47.91,
   "UD-Q3_K_XL": 51.02,
   "UD-Q4_K_XL": 63.07,
   "UD-Q5_K_XL": 77.16,
   "UD-Q6_K_XL": 94.58,
   "UD-Q8_K_XL": 118.97
  }
 },
 "runsWith": [
  "vLLM",
  "SGLang",
  "llama.cpp (Ollama, LM Studio)",
  "MLX (Mac)"
 ],
 "verified": {
  "gguf": {
   "day": "2026-10-14",
   "repo": "unsloth/GLM-4.5-Air-GGUF",
   "kv": "match",
   "ours": 188416,
   "theirs": 188416
  }
 },
 "updated": "2026-10-14T00:07:05.854Z"
}