gpt-oss-20b benchmark on an M4 Max · 36 GB

{
  "runId": "run_1e1ca730-8a3d-4948-ab4a-5aa919c8bd44",
  "bundleId": "mlx-gpt-oss-20b-mlx-8bit-361274",
  "status": "verified",
  "promptTokens": 40960,
  "completionTokens": 10240,
  "contextLength": 5120,
  "harness": {
    "version": "0.1.20",
    "gitSha": "unknown"
  },
  "runtime": {
    "name": "mlx_lm",
    "version": "0.31.2",
    "buildFlags": null
  },
  "model": {
    "displayName": "gpt-oss-20b",
    "format": "mlx",
    "quant": "8bit",
    "architecture": "gpt_oss",
    "source": "lmstudio-community/gpt-oss-20b-MLX-8bit",
    "fileSizeBytes": 22228562304,
    "lab": {
      "name": "OpenAI",
      "slug": "openai"
    },
    "quantizedBy": {
      "name": "LM Studio",
      "slug": "lmstudio"
    }
  },
  "device": {
    "cpu": "Apple M4 Max",
    "cpuCores": 14,
    "gpu": "Apple M4 Max",
    "gpuCores": 32,
    "gpuCount": 1,
    "ramGb": 36,
    "osName": "macOS",
    "osVersion": "15.6.1"
  },
  "decodeTpsMean": 69.8,
  "prefillTpsMean": 1302.4,
  "ttftP50Ms": 3144.83,
  "idleTpsMean": 427,
  "peakRssMb": 22528,
  "trialsPassed": 10,
  "trialsTotal": 10,
  "runnabilityScore": 0.6079152380952381,
  "bundleSha256": "cc9dcbb51b80512f9060b2c6b01134ac978ca9c6dbf3a4127a504395f0405eac",
  "createdAt": "2026-04-11T02:31:47.938Z"
}