Qwen3.6-35B-A3B benchmark on an M4 Pro · 48 GB

{
  "runId": "run_a497f5c5-fa2c-4829-8f34-3af6e187a212",
  "bundleId": "mlx-qwen3.6-35b-a3b-8bit-38026b",
  "status": "verified",
  "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": "Qwen3.6-35B-A3B",
    "format": "mlx",
    "quant": "8bit",
    "architecture": "qwen3_5_moe",
    "source": "mlx-community/Qwen3.6-35B-A3B-8bit",
    "fileSizeBytes": 37721403002,
    "lab": {
      "name": "Qwen",
      "slug": "qwen"
    },
    "quantizedBy": {
      "name": "MLX Community",
      "slug": "mlx-community"
    }
  },
  "device": {
    "cpu": "Apple M4 Pro",
    "cpuCores": 14,
    "gpu": "Apple M4 Pro",
    "gpuCores": 20,
    "gpuCount": 1,
    "ramGb": 48,
    "osName": "macOS",
    "osVersion": "26.5.2"
  },
  "decodeTpsMean": 51.5,
  "prefillTpsMean": 873.3,
  "ttftP50Ms": 4692.02,
  "idleTpsMean": 6546,
  "peakRssMb": 37888,
  "trialsPassed": 10,
  "trialsTotal": 10,
  "runnabilityScore": 0.51458,
  "bundleSha256": "88f4bb751cb62a7eb4c355c6b6483b5c3128f0fed42dfe77bd60de653e0dc17e",
  "createdAt": "2026-07-02T05:52:06.527Z"
}