gpt-oss-20b benchmark on an M4 Max · 128 GB

{
  "runId": "run_a1fc0605-6359-4a13-882a-0b2eee267c20",
  "bundleId": "mlx-gpt-oss-20b-df6376",
  "status": "verified",
  "promptTokens": 40960,
  "completionTokens": 10240,
  "contextLength": 5120,
  "harness": {
    "version": "0.1.18",
    "gitSha": "c60e6d1"
  },
  "runtime": {
    "name": "mlx_lm",
    "version": "0.31.2",
    "buildFlags": null
  },
  "model": {
    "displayName": "gpt-oss-20b",
    "format": "mlx",
    "quant": "4bit",
    "architecture": "gpt_oss",
    "source": "openai/gpt-oss-20b",
    "fileSizeBytes": 13761316904,
    "lab": {
      "name": "OpenAI",
      "slug": "openai"
    },
    "quantizedBy": {
      "name": "OpenAI",
      "slug": "openai"
    }
  },
  "device": {
    "cpu": "Apple M4 Max",
    "cpuCores": 16,
    "gpu": "Apple M4 Max",
    "gpuCores": 40,
    "gpuCount": 1,
    "ramGb": 128,
    "osName": "macOS",
    "osVersion": "26.4.1"
  },
  "decodeTpsMean": 91.3,
  "prefillTpsMean": 1307.2,
  "ttftP50Ms": 3153.42,
  "idleTpsMean": 4164,
  "peakRssMb": 18432,
  "trialsPassed": 10,
  "trialsTotal": 10,
  "runnabilityScore": 0.842042142857143,
  "bundleSha256": "673ad9ce5a166cfc9f0d6726be283f47c158f549c1fe96b5b60edc8b00068768",
  "createdAt": "2026-04-24T19:35:12.581Z"
}