Gemma 4 31B IT benchmark on an M5 Pro · 48 GB

{
  "runId": "run_d4927794-ebb9-4603-93ce-dfefaff3e572",
  "bundleId": "llamacpp-gemma-4-31b-it-q4_k_m.gguf-835d56",
  "status": "verified",
  "promptTokens": 40960,
  "completionTokens": 10240,
  "contextLength": 5120,
  "harness": {
    "version": "0.1.20",
    "gitSha": "unknown"
  },
  "runtime": {
    "name": "llama.cpp",
    "version": "b8680",
    "buildFlags": "metal"
  },
  "model": {
    "displayName": "Gemma 4 31B IT",
    "format": "gguf",
    "quant": "q4_k_m",
    "architecture": "gemma4",
    "source": "unsloth/gemma-4-31B-it-GGUF:gemma-4-31B-it-Q4_K_M.gguf",
    "fileSizeBytes": 18323730976,
    "lab": {
      "name": "Google",
      "slug": "google"
    },
    "quantizedBy": {
      "name": "Unsloth",
      "slug": "unsloth"
    }
  },
  "device": {
    "cpu": "Apple M5 Pro",
    "cpuCores": 15,
    "gpu": "Apple M5 Pro",
    "gpuCores": 16,
    "gpuCount": 1,
    "ramGb": 48,
    "osName": "macOS",
    "osVersion": "26.4.1"
  },
  "decodeTpsMean": 13,
  "prefillTpsMean": 191.9,
  "ttftP50Ms": 21123.09,
  "idleTpsMean": 7226,
  "peakRssMb": 10883.5,
  "trialsPassed": 10,
  "trialsTotal": 10,
  "runnabilityScore": 0.46007841099330354,
  "bundleSha256": "913ac9d82efa86b26680be26fac2b86f2d0de7ef89e775f12fb70f94503fd138",
  "createdAt": "2026-04-13T17:43:31.860Z"
}