{
 "id": "spark-x2.5-4b",
 "name": "Spark-X2.5 4B",
 "repo": "XHToken/Spark-X2.5-4B",
 "lab": {
  "name": "iFlytek",
  "country": "China"
 },
 "released": "2026-08-24",
 "parameters": 4.11207936,
 "activeParameters": null,
 "maxContext": 1048576,
 "architecture": "36 layers · 9 full attention (4 KV × 256) · 27 sliding-window layers (512 tokens, 4 KV × 256)",
 "page": "https://studiotvai.com/vram-requirements/spark-x2-5-4b",
 "calculator": "https://studiotvai.com/llm-vram-calculator?model=spark-x2.5-4b",
 "weightsGiB": {
  "bf16": 7.66,
  "fp8": 4.14,
  "q4_k_m": 2.38
 },
 "kvCacheGiBPerRequest": [
  {
   "context": 8192,
   "bf16": 0.334,
   "fp8": 0.167
  },
  {
   "context": 32768,
   "bf16": 1.178,
   "fp8": 0.589
  },
  {
   "context": 131072,
   "bf16": 4.553,
   "fp8": 2.276
  },
  {
   "context": 1048576,
   "bf16": 36.053,
   "fp8": 18.026
  }
 ],
 "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": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "rtx-5090",
    "name": "GeForce RTX 5090",
    "memoryGB": 32,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 1,
     "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": "abenzerps/Spark-X2.5-4B-GGUF",
  "gib": {
   "Q4_0": 2.24,
   "Q4_K_M": 2.42,
   "Q5_K_M": 2.77,
   "Q6_K": 3.15,
   "Q8_0": 4.07
  }
 },
 "runsWith": [
  "vLLM",
  "SGLang",
  "llama.cpp (Ollama, LM Studio)",
  "MLX (Mac)"
 ],
 "verified": null,
 "updated": "2026-10-10T15:05:26.095Z"
}