{
 "id": "clef",
 "name": "Clef",
 "repo": "Cloudflare/clef",
 "lab": {
  "name": "Cloudflare",
  "country": "United States"
 },
 "released": "2026-09-30",
 "parameters": 27.35672856,
 "activeParameters": null,
 "maxContext": 262144,
 "architecture": "64 layers · 16 full attention (4 KV × 256) · 48 linear attention",
 "page": "https://studiotvai.com/vram-requirements/clef",
 "calculator": "https://studiotvai.com/llm-vram-calculator?model=clef",
 "weightsGiB": {
  "bf16": 50.96,
  "fp8": 27.85,
  "q4_k_m": 15.38
 },
 "kvCacheGiBPerRequest": [
  {
   "context": 8192,
   "bf16": 0.5,
   "fp8": 0.25
  },
  {
   "context": 32768,
   "bf16": 2,
   "fp8": 1
  },
  {
   "context": 131072,
   "bf16": 8,
   "fp8": 4
  },
  {
   "context": 262144,
   "bf16": 16,
   "fp8": 8
  }
 ],
 "recurrentStateGiBPerRequest": 0.1434,
 "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": "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": "bartowski/Cloudflare_clef-GGUF",
  "gib": {
   "IQ3_M": 12.83,
   "IQ4_NL": 15.18,
   "IQ4_XS": 14.49,
   "Q2_K": 10.9,
   "Q3_K_L": 14.11,
   "Q3_K_M": 13.48,
   "Q3_K_S": 12.66,
   "Q4_0": 15.07,
   "Q4_1": 16.49,
   "Q4_K_L": 17.32,
   "Q4_K_M": 16.44,
   "Q4_K_S": 15.45,
   "Q5_K_M": 19.22,
   "Q5_K_S": 18.23,
   "Q6_K": 21.74,
   "Q6_K_L": 22.31,
   "Q8_0": 26.82,
   "BF16": 50.35
  }
 },
 "runsWith": [
  "vLLM",
  "SGLang",
  "llama.cpp (Ollama, LM Studio)",
  "MLX (Mac)"
 ],
 "verified": null,
 "updated": "2026-10-10T15:05:26.095Z"
}