{
 "id": "inkling-small",
 "name": "Inkling Small",
 "repo": "thinkingmachines/Inkling-Small",
 "lab": {
  "name": "Thinking Machines Lab",
  "country": "United States"
 },
 "released": "2026-07-27",
 "parameters": 265.96,
 "activeParameters": 12,
 "maxContext": 1048576,
 "architecture": "MoE 256 experts + 2 shared · 7 global layers + 35 sliding-window layers (512 tokens), GQA 8 KV × 128 · short convolutions · image and audio input",
 "page": "https://studiotvai.com/vram-requirements/inkling-small",
 "calculator": "https://studiotvai.com/llm-vram-calculator?model=inkling-small",
 "weightsGiB": {
  "native": 154.85,
  "bf16": 491.23,
  "fp8": 247.15,
  "q4_k_m": 148.42
 },
 "kvCacheGiBPerRequest": [
  {
   "context": 8192,
   "bf16": 0.287,
   "fp8": 0.144
  },
  {
   "context": 32768,
   "bf16": 0.943,
   "fp8": 0.472
  },
  {
   "context": 131072,
   "bf16": 3.568,
   "fp8": 1.784
  },
  {
   "context": 1048576,
   "bf16": 28.068,
   "fp8": 14.034
  }
 ],
 "recurrentStateGiBPerRequest": 0.0051,
 "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": 8,
     "bf16": 24,
     "fp8": 16,
     "q4_k_m": 7
    }
   },
   {
    "gpu": "rtx-5090",
    "name": "GeForce RTX 5090",
    "memoryGB": 32,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 6,
     "bf16": 24,
     "fp8": 16,
     "q4_k_m": 6
    }
   },
   {
    "gpu": "rtx-pro-6000",
    "name": "RTX PRO 6000 Blackwell",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 2,
     "bf16": 6,
     "fp8": 3,
     "q4_k_m": 2
    }
   },
   {
    "gpu": "mac-m5-max-128",
    "name": "Mac M5 Max 128 GB",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 2,
     "bf16": 6,
     "fp8": 3,
     "q4_k_m": 2
    }
   },
   {
    "gpu": "strix-halo",
    "name": "Ryzen AI Max+ 395 128 GB",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 2,
     "bf16": 6,
     "fp8": 3,
     "q4_k_m": 2
    }
   },
   {
    "gpu": "dgx-spark",
    "name": "DGX Spark (GB10)",
    "memoryGB": 128,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 2,
     "bf16": 4,
     "fp8": 2,
     "q4_k_m": 2
    }
   },
   {
    "gpu": "h100",
    "name": "NVIDIA H100",
    "memoryGB": 80,
    "engine": "vLLM",
    "byFormat": {
     "native": 3,
     "bf16": 8,
     "fp8": 4,
     "q4_k_m": 3
    }
   },
   {
    "gpu": "h200",
    "name": "NVIDIA H200",
    "memoryGB": 141,
    "engine": "vLLM",
    "byFormat": {
     "native": 2,
     "bf16": 5,
     "fp8": 3,
     "q4_k_m": 2
    }
   },
   {
    "gpu": "b200",
    "name": "NVIDIA B200 (HGX)",
    "memoryGB": 180,
    "engine": "vLLM",
    "byFormat": {
     "native": 1,
     "bf16": 4,
     "fp8": 2,
     "q4_k_m": 1
    }
   }
  ]
 },
 "ggufFiles": {
  "repo": "unsloth/Inkling-Small-GGUF",
  "gib": {
   "BF16": 491.19,
   "MXFP4_MOE": 147.19,
   "Q8_0": 261.03,
   "UD-IQ1_M": 73.42,
   "UD-IQ1_S": 69.64,
   "UD-IQ2_M": 76.78,
   "UD-IQ2_XXS": 76.64,
   "UD-IQ3_S": 100.08,
   "UD-IQ3_XXS": 91.21,
   "UD-IQ4_NL": 121.16,
   "UD-IQ4_XS": 118.66,
   "UD-Q2_K_XL": 81.9,
   "UD-Q3_K_M": 111.16,
   "UD-Q3_K_XL": 111.34,
   "UD-Q4_K_M": 151.37,
   "UD-Q4_K_S": 141.87,
   "UD-Q4_K_XL": 152.06,
   "UD-Q5_K_M": 182.62,
   "UD-Q5_K_S": 171.56,
   "UD-Q5_K_XL": 183.15,
   "UD-Q6_K": 203.87,
   "UD-Q6_K_XL": 223.25,
   "UD-Q8_K_XL": 268.61
  }
 },
 "runsWith": [
  "vLLM",
  "SGLang",
  "MLX (Mac)"
 ],
 "verified": null,
 "updated": "2026-10-30T00:07:05.915Z"
}