{
 "id": "longcat-2.5",
 "name": "LongCat-2.5",
 "repo": "meituan-longcat/LongCat-2.5",
 "lab": {
  "name": "Meituan (LongCat)",
  "country": "China"
 },
 "released": "2026-11-04",
 "parameters": 1779.75,
 "activeParameters": 48,
 "maxContext": 524288,
 "architecture": "MoE 768 experts + 128 zero experts · 2 MLA attentions per layer: 76 caches (512 + 64 latent), sparse-attention indexer on 38 · 135B n-gram embeddings · image and video input",
 "page": "https://studiotvai.com/vram-requirements/longcat-2-5",
 "calculator": "https://studiotvai.com/llm-vram-calculator?model=longcat-2.5",
 "weightsGiB": {
  "bf16": 3057.97,
  "fp8": 1531.48,
  "q4_k_m": 921.64
 },
 "kvCacheGiBPerRequest": [
  {
   "context": 8192,
   "bf16": 0.706,
   "fp8": 0.419
  },
  {
   "context": 32768,
   "bf16": 2.825,
   "fp8": 1.675
  },
  {
   "context": 131072,
   "bf16": 11.3,
   "fp8": 6.698
  },
  {
   "context": 524288,
   "bf16": 45.199,
   "fp8": 26.793
  }
 ],
 "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": null,
     "fp8": 96,
     "q4_k_m": 48
    }
   },
   {
    "gpu": "rtx-5090",
    "name": "GeForce RTX 5090",
    "memoryGB": 32,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 128,
     "fp8": 64,
     "q4_k_m": 32
    }
   },
   {
    "gpu": "rtx-pro-6000",
    "name": "RTX PRO 6000 Blackwell",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 32,
     "fp8": 16,
     "q4_k_m": 16
    }
   },
   {
    "gpu": "mac-m5-max-128",
    "name": "Mac M5 Max 128 GB",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 32,
     "fp8": 16,
     "q4_k_m": 16
    }
   },
   {
    "gpu": "strix-halo",
    "name": "Ryzen AI Max+ 395 128 GB",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 32,
     "fp8": 16,
     "q4_k_m": 16
    }
   },
   {
    "gpu": "dgx-spark",
    "name": "DGX Spark (GB10)",
    "memoryGB": 128,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 24,
     "fp8": 16,
     "q4_k_m": 7
    }
   },
   {
    "gpu": "h100",
    "name": "NVIDIA H100",
    "memoryGB": 80,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 48,
     "fp8": 24,
     "q4_k_m": 16
    }
   },
   {
    "gpu": "h200",
    "name": "NVIDIA H200",
    "memoryGB": 141,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 32,
     "fp8": 16,
     "q4_k_m": 8
    }
   },
   {
    "gpu": "b200",
    "name": "NVIDIA B200 (HGX)",
    "memoryGB": 180,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 24,
     "fp8": 16,
     "q4_k_m": 6
    }
   }
  ]
 },
 "ggufFiles": null,
 "runsWith": [],
 "verified": null,
 "updated": "2026-11-22T00:05:05.824Z"
}
