Gemma 4 E4B IT benchmark on an M3 Max · 36 GB

{
  "runId": "run_c3d224ff-5bf2-46b6-9098-9db8c789b232",
  "bundleId": "mlx-gemma-4-e2b-it-4bit-8557a5",
  "status": "verified",
  "promptTokens": 40960,
  "completionTokens": 10240,
  "contextLength": 5120,
  "harness": {
    "version": "0.1.21",
    "gitSha": "e7fabcf0"
  },
  "runtime": {
    "name": "mlx_lm",
    "version": "0.31.3",
    "buildFlags": null
  },
  "model": {
    "displayName": "Gemma 4 E4B IT",
    "format": "mlx",
    "quant": "4bit",
    "architecture": "gemma4",
    "source": "mlx-community/gemma-4-e2b-it-4bit",
    "fileSizeBytes": 3550670554,
    "lab": {
      "name": "Google",
      "slug": "google"
    },
    "quantizedBy": {
      "name": "MLX Community",
      "slug": "mlx-community"
    }
  },
  "device": {
    "cpu": "Apple M3 Max",
    "cpuCores": 14,
    "gpu": "Apple M3 Max",
    "gpuCores": 30,
    "gpuCount": 1,
    "ramGb": 36,
    "osName": "macOS",
    "osVersion": "26.5"
  },
  "decodeTpsMean": 106.4,
  "prefillTpsMean": 4184.5,
  "ttftP50Ms": 980.85,
  "idleTpsMean": 2338,
  "peakRssMb": 6063,
  "trialsPassed": 10,
  "trialsTotal": 10,
  "runnabilityScore": 0.9295131138392857,
  "bundleSha256": "fb721f38f87cf7806a6a2a68b6a2325bb5ceaf33e39541ff6fea609bc4a42f10",
  "createdAt": "2026-08-05T05:33:41.625Z"
}