Qwen3.6-27B benchmark on an M5 Max · 128 GB

{
  "runId": "run_44426b90-d1be-4be5-98c7-0bc3e6ff9209",
  "bundleId": "mlx-qwen3.6-27b-mlx-8bit-9d94a3",
  "status": "pending",
  "promptTokens": 40960,
  "completionTokens": 10240,
  "contextLength": 5120,
  "harness": {
    "version": "0.1.21",
    "gitSha": "unknown"
  },
  "runtime": {
    "name": "mlx_lm",
    "version": "0.31.3",
    "buildFlags": null
  },
  "model": {
    "displayName": "Qwen3.6-27B",
    "format": "mlx",
    "quant": "8bit",
    "architecture": "qwen3_5",
    "source": "unsloth/Qwen3.6-27B-MLX-8bit",
    "fileSizeBytes": 34713723613,
    "lab": {
      "name": "Qwen",
      "slug": "qwen"
    },
    "quantizedBy": {
      "name": "Unsloth",
      "slug": "unsloth"
    }
  },
  "device": {
    "cpu": "Apple M5 Max",
    "cpuCores": 18,
    "gpu": "Apple M5 Max",
    "gpuCores": 40,
    "gpuCount": 1,
    "ramGb": 128,
    "osName": "macOS",
    "osVersion": "26.5"
  },
  "decodeTpsMean": 14.8,
  "prefillTpsMean": 509.7,
  "ttftP50Ms": 8102.47,
  "idleTpsMean": 2428,
  "peakRssMb": 36864,
  "trialsPassed": 10,
  "trialsTotal": 10,
  "runnabilityScore": 0.5036342857142858,
  "bundleSha256": "97f7ec0b3660c6e829d0c219288aa67bd26f03568cc0c62e36c5b88862bc1585",
  "createdAt": "2026-06-02T04:40:15.903Z"
}