{
 "id": "inkling-2-mini",
 "name": "Inkling 2 Mini",
 "repo": "thinkingmachines/Inkling-2-Mini",
 "lab": {
  "name": "Thinking Machines Lab",
  "country": "United States"
 },
 "released": "2026-11-20",
 "parameters": 26.39,
 "activeParameters": 3.4,
 "maxContext": 1048576,
 "architecture": "Looped MoE: 24 layers, the middle 12 run 3 times (48 layer visits, each with its own cache) · 128 experts · 8 global + 40 sliding-window visits (1024 tokens), GQA 8 KV × 128 · short convolutions · image input",
 "page": "https://studiotvai.com/vram-requirements/inkling-2-mini",
 "calculator": "https://studiotvai.com/llm-vram-calculator?model=inkling-2-mini",
 "weightsGiB": {
  "native": 18.19,
  "bf16": 49.16,
  "fp8": 25.44,
  "q4_k_m": 15.85
 },
 "kvCacheGiBPerRequest": [
  {
   "context": 8192,
   "bf16": 0.406,
   "fp8": 0.203
  },
  {
   "context": 32768,
   "bf16": 1.156,
   "fp8": 0.578
  },
  {
   "context": 131072,
   "bf16": 4.156,
   "fp8": 2.078
  },
  {
   "context": 1048576,
   "bf16": 32.156,
   "fp8": 16.078
  }
 ],
 "recurrentStateGiBPerRequest": 0.0029,
 "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": 1,
     "bf16": 3,
     "fp8": 2,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "rtx-5090",
    "name": "GeForce RTX 5090",
    "memoryGB": 32,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 1,
     "bf16": 2,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "rtx-pro-6000",
    "name": "RTX PRO 6000 Blackwell",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 1,
     "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": {
     "native": 1,
     "bf16": 1,
     "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": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "dgx-spark",
    "name": "DGX Spark (GB10)",
    "memoryGB": 128,
    "engine": "llama.cpp",
    "byFormat": {
     "native": 1,
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "h100",
    "name": "NVIDIA H100",
    "memoryGB": 80,
    "engine": "vLLM",
    "byFormat": {
     "native": 1,
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "h200",
    "name": "NVIDIA H200",
    "memoryGB": 141,
    "engine": "vLLM",
    "byFormat": {
     "native": 1,
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "b200",
    "name": "NVIDIA B200 (HGX)",
    "memoryGB": 180,
    "engine": "vLLM",
    "byFormat": {
     "native": 1,
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   }
  ]
 },
 "ggufFiles": {
  "repo": "AtomicChat/Inkling-2-Mini-GGUF",
  "gib": {
   "BF16": 48.64,
   "IQ1_M": 6.76,
   "IQ1_S": 6.41,
   "IQ2_M": 8.07,
   "IQ2_S": 10.76,
   "Q2": 7.36,
   "IQ2_XXS": 9.71,
   "IQ3_S": 12.42,
   "IQ3_XXS": 11.71,
   "IQ4_XS": 13.97,
   "Q4_K": 15.39,
   "Q5_K": 17.76,
   "Q6_K": 21.8,
   "Q4_K_M": 16.19,
   "Q5_K_M": 18.45,
   "Q8_0": 25.83
  }
 },
 "runsWith": [
  "vLLM",
  "llama.cpp (Ollama, LM Studio)",
  "MLX (Mac)"
 ],
 "verified": null,
 "updated": "2026-11-22T00:05:05.824Z"
}
