{
  "evidenceVersion": "local-ollama-speed-v1",
  "capturedAt": "2026-06-01T21:55:25.064Z",
  "hardwareSlug": "rtx-3080-16gb-mobile",
  "modelTag": "qwen2.5:7b",
  "catalogSlug": "qwen-2.5-7b-instruct",
  "quant": "Q4_K_M",
  "contextSize": 4096,
  "numPredict": 256,
  "prompt": "Write a detailed explanation of how transformer attention works.",
  "scenario": "single-stream",
  "runtime": "ollama-api",
  "command": "POST http://localhost:11434/api/generate model=qwen2.5:7b temperature=0 top_p=1 seed=42 num_ctx=4096 num_predict=256",
  "env": {
    "os": "Microsoft Windows [Version 10.0.26200.8457]",
    "platform": "win32",
    "arch": "x64",
    "cpu": "AMD Ryzen 9 5900HX with Radeon Graphics        ",
    "cpuThreads": 16,
    "ramGb": 31.9,
    "gpu": {
      "vendor": "nvidia",
      "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
      "driver": "571.96",
      "memoryTotalMb": 16384,
      "memoryUsedMb": 1873,
      "tempC": 61,
      "smClockMhz": 210,
      "powerDrawW": 17.85
    },
    "ollamaVersionCli": "ollama version is 0.24.0",
    "ollamaApiVersion": {
      "version": "0.24.0"
    }
  },
  "ollamaTagDetails": {
    "name": "qwen2.5:7b",
    "model": "qwen2.5:7b",
    "modified_at": "2026-06-02T00:54:56.3443284+03:00",
    "size": 4683087332,
    "digest": "845dbda0ea48ed749caafd9e6037047aa19acfcfd82e704d7ca97d631a0b697e",
    "details": {
      "parent_model": "",
      "format": "gguf",
      "family": "qwen2",
      "families": [
        "qwen2"
      ],
      "parameter_size": "7.6B",
      "quantization_level": "Q4_K_M"
    }
  },
  "notes": null,
  "logHashSha256": "0a3b51c3ef429fd8fd597cd7475ab6f28bf7d89549a51ace0a0b91f1327a46f7",
  "runsCaptured": 5,
  "decodeStatsAllRuns": {
    "min": 78.92496064465413,
    "max": 81.75428602433155,
    "median": 80.65176709281884,
    "p5": 79.06043038573516,
    "p95": 81.64930044046277,
    "mean": 80.43253624337028,
    "n": 5
  },
  "decodeStatsSteadyRuns": {
    "min": 78.92496064465413,
    "max": 81.75428602433155,
    "median": 80.41583372752345,
    "p5": 79.0265629504649,
    "p95": 81.67554683642996,
    "mean": 80.37772853100813,
    "n": 4
  },
  "runs": [
    {
      "runIndex": 1,
      "startedAt": "2026-06-01T21:55:00.512Z",
      "finishedAt": "2026-06-01T21:55:07.698Z",
      "rawPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/qwen2-5-7b/run1.ollama.json",
      "logPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/qwen2-5-7b/run1.log",
      "decodeTokPerSec": 80.65176709281884,
      "prefillTokPerSec": 971.9747286570549,
      "totalMs": 7098.7584,
      "loadMs": 3252.4522,
      "promptEvalMs": 40.1245,
      "evalMs": 3174.14,
      "promptTokens": 39,
      "responseTokens": 256,
      "responseHashSha256": "bd186d77498ff12b1e33edc042e0dccf2c38964d44b2927655406331aa282945",
      "beforeGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 1873,
        "tempC": 61,
        "smClockMhz": 210,
        "powerDrawW": 18.43
      },
      "afterGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 6818,
        "tempC": 67,
        "smClockMhz": 1230,
        "powerDrawW": 114.61
      }
    },
    {
      "runIndex": 2,
      "startedAt": "2026-06-01T21:55:07.779Z",
      "finishedAt": "2026-06-01T21:55:12.057Z",
      "rawPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/qwen2-5-7b/run2.ollama.json",
      "logPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/qwen2-5-7b/run2.log",
      "decodeTokPerSec": 81.75428602433155,
      "prefillTokPerSec": 2592.9127052722556,
      "totalMs": 4170.3202,
      "loadMs": 338.2203,
      "promptEvalMs": 15.041,
      "evalMs": 3131.3343,
      "promptTokens": 39,
      "responseTokens": 256,
      "responseHashSha256": "81db393da6491633afc2c0a8afeccc01fbe486e0caaef8f4877f745613210a3a",
      "beforeGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 6818,
        "tempC": 67,
        "smClockMhz": 1605,
        "powerDrawW": 108.59
      },
      "afterGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 6818,
        "tempC": 68,
        "smClockMhz": 1230,
        "powerDrawW": 111.43
      }
    },
    {
      "runIndex": 3,
      "startedAt": "2026-06-01T21:55:12.137Z",
      "finishedAt": "2026-06-01T21:55:16.420Z",
      "rawPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/qwen2-5-7b/run3.ollama.json",
      "logPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/qwen2-5-7b/run3.log",
      "decodeTokPerSec": 79.60230935005927,
      "prefillTokPerSec": 2642.8493982435216,
      "totalMs": 4191.8666,
      "loadMs": 348.7835,
      "promptEvalMs": 14.7568,
      "evalMs": 3215.9871,
      "promptTokens": 39,
      "responseTokens": 256,
      "responseHashSha256": "81db393da6491633afc2c0a8afeccc01fbe486e0caaef8f4877f745613210a3a",
      "beforeGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 6818,
        "tempC": 67,
        "smClockMhz": 1230,
        "powerDrawW": 105.16
      },
      "afterGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 6818,
        "tempC": 70,
        "smClockMhz": 1140,
        "powerDrawW": 110.03
      }
    },
    {
      "runIndex": 4,
      "startedAt": "2026-06-01T21:55:16.496Z",
      "finishedAt": "2026-06-01T21:55:20.731Z",
      "rawPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/qwen2-5-7b/run4.ollama.json",
      "logPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/qwen2-5-7b/run4.log",
      "decodeTokPerSec": 81.22935810498761,
      "prefillTokPerSec": 2466.0756516130664,
      "totalMs": 4143.6463,
      "loadMs": 289.7131,
      "promptEvalMs": 15.8146,
      "evalMs": 3151.5699,
      "promptTokens": 39,
      "responseTokens": 256,
      "responseHashSha256": "81db393da6491633afc2c0a8afeccc01fbe486e0caaef8f4877f745613210a3a",
      "beforeGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 6818,
        "tempC": 70,
        "smClockMhz": 1140,
        "powerDrawW": 110.03
      },
      "afterGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 6818,
        "tempC": 71,
        "smClockMhz": 1230,
        "powerDrawW": 110.86
      }
    },
    {
      "runIndex": 5,
      "startedAt": "2026-06-01T21:55:20.807Z",
      "finishedAt": "2026-06-01T21:55:25.062Z",
      "rawPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/qwen2-5-7b/run5.ollama.json",
      "logPath": "public/benchmarks/evidence/local-ollama-rtx-3080-16gb-mobile-2026-06-01/qwen2-5-7b/run5.log",
      "decodeTokPerSec": 78.92496064465413,
      "prefillTokPerSec": 2615.553826756445,
      "totalMs": 4166.819,
      "loadMs": 302.2888,
      "promptEvalMs": 14.9108,
      "evalMs": 3243.5873,
      "promptTokens": 39,
      "responseTokens": 256,
      "responseHashSha256": "81db393da6491633afc2c0a8afeccc01fbe486e0caaef8f4877f745613210a3a",
      "beforeGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 6818,
        "tempC": 71,
        "smClockMhz": 1770,
        "powerDrawW": 110.86
      },
      "afterGpu": {
        "vendor": "nvidia",
        "name": "NVIDIA GeForce RTX 3080 Laptop GPU",
        "driver": "571.96",
        "memoryTotalMb": 16384,
        "memoryUsedMb": 6818,
        "tempC": 72,
        "smClockMhz": 1170,
        "powerDrawW": 113.14
      }
    }
  ],
  "errors": [],
  "uploadEligible": true
}