{
 "id": "glm-4.7-flash",
 "name": "GLM-4.7-Flash",
 "repo": "zai-org/GLM-4.7-Flash",
 "lab": {
  "name": "Z.ai",
  "country": "China"
 },
 "released": "2026-01-19",
 "parameters": 31.221488576,
 "activeParameters": 3.6,
 "maxContext": 202752,
 "architecture": "47 layers · MLA (512 + 64 latent) on 47",
 "page": "https://studiotvai.com/vram-requirements/glm-4-7-flash",
 "calculator": "https://studiotvai.com/llm-vram-calculator?model=glm-4.7-flash",
 "weightsGiB": {
  "bf16": 55.77,
  "fp8": 28.48,
  "q4_k_m": 16.9
 },
 "kvCacheGiBPerRequest": [
  {
   "context": 8192,
   "bf16": 0.413,
   "fp8": 0.207
  },
  {
   "context": 32768,
   "bf16": 1.652,
   "fp8": 0.826
  },
  {
   "context": 131072,
   "bf16": 6.609,
   "fp8": 3.305
  },
  {
   "context": 202752,
   "bf16": 10.224,
   "fp8": 5.112
  }
 ],
 "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": 3,
     "fp8": 2,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "rtx-5090",
    "name": "GeForce RTX 5090",
    "memoryGB": 32,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 2,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "rtx-pro-6000",
    "name": "RTX PRO 6000 Blackwell",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "mac-m5-max-128",
    "name": "Mac M5 Max 128 GB",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "strix-halo",
    "name": "Ryzen AI Max+ 395 128 GB",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "dgx-spark",
    "name": "DGX Spark (GB10)",
    "memoryGB": 128,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "h100",
    "name": "NVIDIA H100",
    "memoryGB": 80,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "h200",
    "name": "NVIDIA H200",
    "memoryGB": 141,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "b200",
    "name": "NVIDIA B200 (HGX)",
    "memoryGB": 180,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   }
  ]
 },
 "ggufFiles": {
  "repo": "unsloth/GLM-4.7-Flash-GGUF",
  "gib": {
   "BF16": 55.79,
   "IQ4_NL": 15.99,
   "IQ4_XS": 15.15,
   "MXFP4_MOE": 15.8,
   "Q2_K": 10.57,
   "Q2_K_L": 10.63,
   "Q3_K_M": 13.61,
   "Q3_K_S": 12.38,
   "Q4_0": 16.03,
   "Q4_1": 17.67,
   "Q4_K_M": 17.05,
   "Q4_K_S": 16.08,
   "Q5_K_M": 19.94,
   "Q5_K_S": 19.39,
   "Q6_K": 23,
   "Q8_0": 29.66,
   "UD-IQ1_M": 9.13,
   "UD-IQ1_S": 8.61,
   "UD-IQ2_M": 10.24,
   "UD-IQ2_XXS": 9.79,
   "UD-IQ3_XXS": 12.02,
   "UD-Q2_K_XL": 11.07,
   "UD-Q3_K_XL": 12.84,
   "UD-Q4_K_XL": 16.32,
   "UD-Q5_K_XL": 20.2,
   "UD-Q6_K_XL": 24.38,
   "UD-Q8_K_XL": 33.18,
   "UD-TQ1_0": 7.76
  }
 },
 "runsWith": [
  "vLLM",
  "SGLang",
  "llama.cpp (Ollama, LM Studio)",
  "MLX (Mac)"
 ],
 "verified": null,
 "updated": "2026-11-06T21:02:26.190Z"
}