Qwen3.5-4B benchmark on an M4 Pro · 24 GB

{
  "runId": "run_01f94cdb-c915-45e4-8951-6cbffa5fded1",
  "bundleId": "mlx-qwen3.5-4b-6bit-3ea216",
  "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": "Qwen3.5-4B",
    "format": "mlx",
    "quant": "6bit",
    "architecture": "qwen3_5",
    "source": "mlx-community/Qwen3.5-4B-6bit",
    "fileSizeBytes": 4085498443,
    "lab": {
      "name": "Qwen",
      "slug": "qwen"
    },
    "quantizedBy": {
      "name": "MLX Community",
      "slug": "mlx-community"
    }
  },
  "device": {
    "cpu": "Apple M4 Pro",
    "cpuCores": 12,
    "gpu": "Apple M4 Pro",
    "gpuCores": 16,
    "gpuCount": 1,
    "ramGb": 24,
    "osName": "macOS",
    "osVersion": "26.6.2"
  },
  "decodeTpsMean": 56.8,
  "prefillTpsMean": 635,
  "ttftP50Ms": 6501.64,
  "idleTpsMean": 2408,
  "peakRssMb": 5678,
  "trialsPassed": 10,
  "trialsTotal": 10,
  "runnabilityScore": 0.6996835379464286,
  "bundleSha256": "5bee327a86d8e1c8f3e47d519dde31d5f3cdf02b18e5fcd14bf84fd2c7dfab3d",
  "createdAt": "2026-09-13T10:41:08.645Z"
}