Qwen3.8-27B benchmark on an M5 Pro · 64 GB

{
  "runId": "run_038d5e98-6d55-46ac-ab78-3a3573e9abef",
  "bundleId": "mlx-qwen3.8-27b-nvfp4-5c2f02",
  "status": "verified",
  "promptTokens": 40960,
  "completionTokens": 10240,
  "contextLength": 5120,
  "harness": {
    "version": "0.1.22",
    "gitSha": "50c24e0"
  },
  "runtime": {
    "name": "mlx_lm",
    "version": "0.31.3",
    "buildFlags": null
  },
  "model": {
    "displayName": "Qwen3.8-27B",
    "format": "mlx",
    "quant": "nvfp4",
    "architecture": "qwen3_5",
    "source": "mlx-community/Qwen3.8-27B-nvfp4",
    "fileSizeBytes": 16054475404,
    "lab": {
      "name": "Qwen",
      "slug": "qwen"
    },
    "quantizedBy": {
      "name": "MLX Community",
      "slug": "mlx-community"
    }
  },
  "device": {
    "cpu": "Apple M5 Pro",
    "cpuCores": 18,
    "gpu": "Apple M5 Pro",
    "gpuCores": 20,
    "gpuCount": 1,
    "ramGb": 64,
    "osName": "macOS",
    "osVersion": "26.5.2"
  },
  "decodeTpsMean": 17.7,
  "prefillTpsMean": 470.8,
  "ttftP50Ms": 8766.41,
  "idleTpsMean": 2612,
  "peakRssMb": 18432,
  "trialsPassed": 10,
  "trialsTotal": 10,
  "runnabilityScore": 0.5238976190476191,
  "bundleSha256": "31a8be33c436aea6844bbecc165a29ebcc8a4a2936384d3866d367db90aa5f84",
  "createdAt": "2026-08-15T12:48:21.123Z"
}