Qwen3.5-0.8B benchmark on an M2 · 16 GB

{
  "runId": "run_8a5b9c1c-e587-4278-b79b-82f6bfb4158b",
  "bundleId": "mlx-qwen3.5-0.8b-mlx-4bit-cc040c",
  "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": "Qwen3.5-0.8B",
    "format": "mlx",
    "quant": "4bit",
    "architecture": "qwen3_5",
    "source": "mlx-community/Qwen3.5-0.8B-4bit",
    "fileSizeBytes": 625229487,
    "lab": {
      "name": "Qwen",
      "slug": "qwen"
    },
    "quantizedBy": {
      "name": "MLX Community",
      "slug": "mlx-community"
    }
  },
  "device": {
    "cpu": "Apple M2",
    "cpuCores": 8,
    "gpu": "Apple M2",
    "gpuCores": 8,
    "gpuCount": 1,
    "ramGb": 16,
    "osName": "macOS",
    "osVersion": "15.7.2"
  },
  "decodeTpsMean": 115.1,
  "prefillTpsMean": 1098.3,
  "ttftP50Ms": 3567.62,
  "idleTpsMean": 501,
  "peakRssMb": 2037,
  "trialsPassed": 10,
  "trialsTotal": 10,
  "runnabilityScore": 0.8516313085937499,
  "bundleSha256": "966f0c6e8ecac815698c1eda168ab49a3c37a2aa6a3b7f2e8a836f00e9968766",
  "createdAt": "2026-09-02T09:32:37.555Z"
}