Gemma 4 26B A4B IT benchmark on an M5 Pro · 64 GB

{
  "runId": "run_1d1127cf-fd9f-40a4-ac29-b1768b904ee8",
  "bundleId": "llamacpp-gemma-4-26b-a4b-it-mxfp4_moe.gguf-062a3f",
  "status": "verified",
  "promptTokens": 40960,
  "completionTokens": 10240,
  "contextLength": 5120,
  "harness": {
    "version": "0.1.22",
    "gitSha": "50c24e0"
  },
  "runtime": {
    "name": "llama.cpp",
    "version": "b8920",
    "buildFlags": "metal"
  },
  "model": {
    "displayName": "Gemma 4 26B A4B IT",
    "format": "gguf",
    "quant": "mxfp4_moe",
    "architecture": "gemma4",
    "source": "unsloth/gemma-4-26B-A4B-it-GGUF:gemma-4-26B-A4B-it-MXFP4_MOE.gguf",
    "fileSizeBytes": 16551046944,
    "lab": {
      "name": "Google",
      "slug": "google"
    },
    "quantizedBy": {
      "name": "Unsloth",
      "slug": "unsloth"
    }
  },
  "device": {
    "cpu": "Apple M5 Pro",
    "cpuCores": 18,
    "gpu": "Apple M5 Pro",
    "gpuCores": 20,
    "gpuCount": 1,
    "ramGb": 64,
    "osName": "macOS",
    "osVersion": "26.5.1"
  },
  "decodeTpsMean": 60.6,
  "prefillTpsMean": 1542.2,
  "ttftP50Ms": 2654.59,
  "idleTpsMean": 9134.8,
  "peakRssMb": 9370.8,
  "trialsPassed": 10,
  "trialsTotal": 10,
  "runnabilityScore": 0.8067298311941964,
  "bundleSha256": "0021ee7960d4caebc8be65f21bd4cb38efc3d9d43cca4cb529c0b9803bc759ed",
  "createdAt": "2026-06-06T22:15:19.315Z"
}