LFM2-350M benchmark on an M4 · 32 GB

{
  "runId": "run_d8546889-1aa0-49b5-ae37-98960c5cd382",
  "bundleId": "llamacpp-lfm2-350m-q4_k_m.gguf-2a3498",
  "status": "verified",
  "promptTokens": 40960,
  "completionTokens": 10240,
  "contextLength": 5120,
  "harness": {
    "version": "0.1.21",
    "gitSha": "f68adca"
  },
  "runtime": {
    "name": "llama.cpp",
    "version": "b8680",
    "buildFlags": "metal"
  },
  "model": {
    "displayName": "LFM2-350M",
    "format": "gguf",
    "quant": "q4_k_m",
    "architecture": "lfm2",
    "source": "LiquidAI/LFM2-350M-GGUF:LFM2-350M-Q4_K_M.gguf",
    "fileSizeBytes": 229309376,
    "lab": {
      "name": "Liquid AI",
      "slug": "liquid"
    },
    "quantizedBy": {
      "name": "Liquid AI",
      "slug": "liquid"
    }
  },
  "device": {
    "cpu": "Apple M4",
    "cpuCores": 10,
    "gpu": "Apple M4",
    "gpuCores": 10,
    "gpuCount": 1,
    "ramGb": 32,
    "osName": "macOS",
    "osVersion": "15.7.4"
  },
  "decodeTpsMean": 203.6,
  "prefillTpsMean": 2912.7,
  "ttftP50Ms": 1533.9,
  "idleTpsMean": 513.5,
  "peakRssMb": 553.1,
  "trialsPassed": 10,
  "trialsTotal": 10,
  "runnabilityScore": 0.9655835260881697,
  "bundleSha256": "d06433c25313c8171ae12c4430ccec1f88978c6a1a0a80e9ca1f0b6ad7debc65",
  "createdAt": "2026-04-17T17:12:19.173Z"
}