{
 "id": "qwen3.8-flash-next-e384-fp4-v2",
 "name": "Qwen3.8-Flash-Next E384 FP4 V2",
 "repo": "autotrust/Qwen3.8-Flash-Next-E384-FP4-V2",
 "lab": {
  "name": "AutoTrust AI",
  "country": "Singapore"
 },
 "released": "2026-10-26",
 "parameters": 149.17,
 "activeParameters": 6,
 "maxContext": 262144,
 "architecture": "Qwen3.8-Flash-Next pruned to 384 experts · 12 Qwen Sparse Attention layers (GQA, 2 KV × 256, + indexer keys per 4-token block) · 36 Gated DeltaNet layers · 51B n-gram embeddings",
 "page": "https://studiotvai.com/vram-requirements/qwen3-8-flash-next-e384-fp4-v2",
 "calculator": "https://studiotvai.com/llm-vram-calculator?model=qwen3.8-flash-next-e384-fp4-v2",
 "weightsGiB": {
  "native": 78.01,
  "bf16": 274.17,
  "fp8": 138.27,
  "q4_k_m": 86.74
 },
 "kvCacheGiBPerRequest": [
  {
   "context": 8192,
   "bf16": 0.193,
   "fp8": 0.1
  },
  {
   "context": 32768,
   "bf16": 0.773,
   "fp8": 0.398
  },
  {
   "context": 131072,
   "bf16": 3.094,
   "fp8": 1.594
  },
  {
   "context": 262144,
   "bf16": 6.188,
   "fp8": 3.188
  }
 ],
 "recurrentStateGiBPerRequest": 0.1075,
 "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": {
     "native": 3,
     "bf16": 16,
     "fp8": 5,
     "q4_k_m": 3
    }
   },
   {
    "gpu": "rtx-5090",
    "name": "GeForce RTX 5090",
    "memoryGB": 32,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 2,
     "bf16": 7,
     "fp8": 4,
     "q4_k_m": 2
    }
   },
   {
    "gpu": "rtx-pro-6000",
    "name": "RTX PRO 6000 Blackwell",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 1,
     "bf16": 2,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "mac-m5-max-128",
    "name": "Mac M5 Max 128 GB",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 1,
     "bf16": 2,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "strix-halo",
    "name": "Ryzen AI Max+ 395 128 GB",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 1,
     "bf16": 2,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "dgx-spark",
    "name": "DGX Spark (GB10)",
    "memoryGB": 128,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 1,
     "bf16": 2,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "h100",
    "name": "NVIDIA H100",
    "memoryGB": 80,
    "engine": "vLLM",
    "byFormat": {
     "native": 2,
     "bf16": 5,
     "fp8": 3,
     "q4_k_m": 2
    }
   },
   {
    "gpu": "h200",
    "name": "NVIDIA H200",
    "memoryGB": 141,
    "engine": "vLLM",
    "byFormat": {
     "native": 1,
     "bf16": 3,
     "fp8": 2,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "b200",
    "name": "NVIDIA B200 (HGX)",
    "memoryGB": 180,
    "engine": "vLLM",
    "byFormat": {
     "native": 1,
     "bf16": 2,
     "fp8": 1,
     "q4_k_m": 1
    }
   }
  ]
 },
 "ggufFiles": null,
 "runsWith": [],
 "verified": null,
 "updated": "2026-11-06T21:02:26.190Z"
}