A model is still running

A model is still running or downloading. Leaving will stop the current operation. Stay on this page to let it finish.

Skip to content
LocalAI.toolsBeta

Join the waitlist

Get an email when the new apps and model runs launch.

No account needed.

SmolLM3 Desktop

General chat, drafting and short multilingual replies locally. Strict-format output is not guaranteed.

Desktop inferencePrivate · Runs on your computer
About this model

Continue a text or code prefix locally with the pretrained SmolLM3 checkpoint.

Created by
Hugging Face
License
Apache-2.0
Model released
Jul 8, 2025
Source repository created
Jun 19, 2025
Model weights · FP16
6.2 GB

Best for

  • Short prose continuation
  • Short code continuation

Keep in mind

  • The pretrained checkpoint continues literal text. Chat templates, instruction following and tool use are outside this adapter's scope.
  • Use a clear prefix. Ambiguous or incomplete context can produce unexpected, inaccurate or unsafe text.
  • Review all generated text and test code before running it.
  • The adapter limits input plus output to 8192 tokens. Long-context behavior is unverified.
  • Measured quality covers two English sentence prefixes and a short Python expression.

Processed on your deviceYour input is processed locally in the desktop app. Model and runtime files download from external hosts.

Target platforms · verification required

  • macos · aarch64, x86_64
  • windows · x86_64

Recommended configuration

FP16 · mps (verified platform only)

Sources & technical details
Original model
HuggingFaceTB/SmolLM3-3B-Base
Release verification
Verified
Release evidence
The July 8 launch article links this exact pretrained Base checkpoint.
Source updated
Aug 14, 2025
Native model files
HuggingFaceTB/SmolLM3-3B-Base
Files updated
Aug 14, 2025
Artifact revision
d78a42f79198603e614095753484a04c10c2b940

Desktop app

Sample result

SmolLM3 3B Base Desktop

Run SmolLM3 Desktop on your computer

Download the desktop app to load this model, provide input and view results locally.

Public desktop downloads are coming soon.

Device requirements

macOS Apple SiliconCPU / FP32Tested22.4 GiBRAM estimate
macOS Apple SiliconMPS / FP16Tested14.9 GiBRAM estimate
Windows Intel / AMDCPU / FP32Unverified22.4 GiBRAM estimate
macOS Intel / AMDCPU / FP32Unverified22.4 GiBRAM estimate
If your device falls short

MemoryAdvisory only: close other apps, reduce image/tile/token/batch settings or choose a smaller verified model/precision. CPU can need more RAM than GPU.

GPUIf the selected GPU runtime fails, existing CPU fallback may recover when enabled, but CPU has its own memory/runtime needs.

PlatformUse a compatible OS/architecture/runtime or another model. These estimates do not add hardware detection or block runs.

Benchmark evidence

Saved desktop measurements apply to the recorded hardware and runtime. Fallback cases report the device that actually ran; they do not prove GPU support on the requested device.

SmolLM3 3B Base Desktop

21 configurations tested
Passed21
Failed0
TestedNov 7, 2026
Runtimetransformers 5.22.1

Quality criteria

No explicit quality gate recorded. Review per-case evidence; full certification is pending.

PrecisionDeviceResultPlatformInference / load
FP16mpsAcceleratedPassedFastmacos · aarch64260 ms / 1654 ms
FP16mpsAcceleratedPassedFastmacos · aarch64264 ms / 1876 ms
FP16mpsAcceleratedPassedFastmacos · aarch64266 ms / 1654 ms
FP16mpsAcceleratedPassedFastmacos · aarch64267 ms / 1876 ms
FP16mpsAcceleratedPassedFastmacos · aarch64274 ms / 1876 ms
FP16mpsAcceleratedPassedFastmacos · aarch64276 ms / 1654 ms
FP16mpsAcceleratedPassedModeratemacos · aarch64448 ms / 2509 ms
FP16mpsAcceleratedPassedModeratemacos · aarch64486 ms / 2509 ms
FP16mpsAcceleratedPassedModeratemacos · aarch64652 ms / 2509 ms
FP32cpuPassedModeratemacos · aarch641141 ms / 1623 ms
FP32cpuPassedModeratemacos · aarch641148 ms / 1623 ms
FP32cpuPassedModeratemacos · aarch641149 ms / 1623 ms
FP32cpuPassedModeratemacos · aarch641152 ms / 1478 ms
FP32cpuPassedModeratemacos · aarch641167 ms / 1507 ms
FP32cpuPassedModeratemacos · aarch641184 ms / 1430 ms
FP32cpuPassedModeratemacos · aarch641188 ms / 1507 ms
FP32cpuPassedModeratemacos · aarch641200 ms / 1478 ms
FP32cpuPassedModeratemacos · aarch641252 ms / 1430 ms
FP32cpuPassedModeratemacos · aarch641259 ms / 1507 ms
FP32cpuPassedModeratemacos · aarch641329 ms / 1430 ms
FP32cpuPassedModeratemacos · aarch641383 ms / 1478 ms
Test setup & sample details
Quality gate
No explicit quality gate recorded. Review the per-case quality evidence; full certification is pending.
Tested
2026-11-07T11:51:08.117204+00:00
Environment
{ "os": "macos", "arch": "aarch64", "hardware": "Apple M3 Max", "osVersion": "27.2" }
Runtime
transformers 5.22.1
Artifact revision
d78a42f79198603e614095753484a04c10c2b940
Fixture
{
  "file": "model-samples/desktop/smollm3/smollm3-3b-base-native-geography.json",
  "size": 152,
  "sha1": "f033f82b9e5c7821b37f56df58bd3f18c26c24ef",
  "sha256": "48f3547630afad495ce68672989f85e41ccbc27ef56df0d2805f36097901dd21",
  "url": "https://assets.localai.tools/samples/v1/files/48f3547630afad495ce68672989f85e41ccbc27ef56df0d2805f36097901dd21/smollm3-3b-base-native-geography.json"
}
Rerun when
Artifact/runtime revisions; Generation/preprocessing changes; Fixture/quality scope; OS/backend behavior
Per-case quality & failure details
[
  {
    "precision": "fp16",
    "device": "mps",
    "status": "pass",
    "quality": {
      "expectedPrefix": " France",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 653.2365409948397,
      "inferenceMs": 651.743542002805,
      "tokensPerSecond": 12.274766813056614,
      "inputTokens": 8,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "mps:0/torch.float16": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "driverAllocatedMemoryBytes": 6504185856,
      "memoryMeasurementScope": "Post-inference driver allocation; not peak",
      "torchVersion": "2.10.0",
      "loadMs": 2509.2017499991925,
      "device": "mps",
      "provider": "Metal/MPS",
      "acceleration": "partial",
      "parameterDtype": "torch.float16",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 3162.4382909940323,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp16",
    "device": "mps",
    "status": "pass",
    "quality": {
      "expectedPrefix": " Celsius",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 486.47320800228044,
      "inferenceMs": 485.66837499674875,
      "tokensPerSecond": 16.47214521648348,
      "inputTokens": 11,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "mps:0/torch.float16": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "driverAllocatedMemoryBytes": 6504267776,
      "memoryMeasurementScope": "Post-inference driver allocation; not peak",
      "torchVersion": "2.10.0",
      "loadMs": 2509.2017499991925,
      "device": "mps",
      "provider": "Metal/MPS",
      "acceleration": "partial",
      "parameterDtype": "torch.float16",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2995.674958001473,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp16",
    "device": "mps",
    "status": "pass",
    "quality": {
      "expectedPrefix": " x",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 449.03729199722875,
      "inferenceMs": 448.14449999830686,
      "tokensPerSecond": 17.851384988614665,
      "inputTokens": 16,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "mps:0/torch.float16": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "driverAllocatedMemoryBytes": 6512754688,
      "memoryMeasurementScope": "Post-inference driver allocation; not peak",
      "torchVersion": "2.10.0",
      "loadMs": 2509.2017499991925,
      "device": "mps",
      "provider": "Metal/MPS",
      "acceleration": "partial",
      "parameterDtype": "torch.float16",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2958.2390419964213,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp16",
    "device": "mps",
    "status": "pass",
    "quality": {
      "expectedPrefix": " France",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 277.80287499626866,
      "inferenceMs": 276.3670420026756,
      "tokensPerSecond": 28.947011705985368,
      "inputTokens": 8,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "mps:0/torch.float16": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "driverAllocatedMemoryBytes": 6504202240,
      "memoryMeasurementScope": "Post-inference driver allocation; not peak",
      "torchVersion": "2.10.0",
      "loadMs": 1654.1304589991341,
      "device": "mps",
      "provider": "Metal/MPS",
      "acceleration": "partial",
      "parameterDtype": "torch.float16",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 1931.9333339954028,
      "grade": "fast",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp16",
    "device": "mps",
    "status": "pass",
    "quality": {
      "expectedPrefix": " Celsius",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 260.59574999817414,
      "inferenceMs": 259.74349999887636,
      "tokensPerSecond": 30.799615774926448,
      "inputTokens": 11,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "mps:0/torch.float16": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "driverAllocatedMemoryBytes": 6504267776,
      "memoryMeasurementScope": "Post-inference driver allocation; not peak",
      "torchVersion": "2.10.0",
      "loadMs": 1654.1304589991341,
      "device": "mps",
      "provider": "Metal/MPS",
      "acceleration": "partial",
      "parameterDtype": "torch.float16",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 1914.7262089973083,
      "grade": "fast",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp16",
    "device": "mps",
    "status": "pass",
    "quality": {
      "expectedPrefix": " x",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 267.21795900084544,
      "inferenceMs": 266.35383399843704,
      "tokensPerSecond": 30.035235010159244,
      "inputTokens": 16,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "mps:0/torch.float16": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "driverAllocatedMemoryBytes": 6512754688,
      "memoryMeasurementScope": "Post-inference driver allocation; not peak",
      "torchVersion": "2.10.0",
      "loadMs": 1654.1304589991341,
      "device": "mps",
      "provider": "Metal/MPS",
      "acceleration": "partial",
      "parameterDtype": "torch.float16",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 1921.3484179999796,
      "grade": "fast",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp16",
    "device": "mps",
    "status": "pass",
    "quality": {
      "expectedPrefix": " France",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 268.126167000446,
      "inferenceMs": 266.76662499812664,
      "tokensPerSecond": 29.98875890136624,
      "inputTokens": 8,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "mps:0/torch.float16": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "driverAllocatedMemoryBytes": 6504202240,
      "memoryMeasurementScope": "Post-inference driver allocation; not peak",
      "torchVersion": "2.10.0",
      "loadMs": 1876.1434170010034,
      "device": "mps",
      "provider": "Metal/MPS",
      "acceleration": "partial",
      "parameterDtype": "torch.float16",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2144.2695840014494,
      "grade": "fast",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp16",
    "device": "mps",
    "status": "pass",
    "quality": {
      "expectedPrefix": " Celsius",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 264.63254199916264,
      "inferenceMs": 263.59941600094317,
      "tokensPerSecond": 30.349080894668507,
      "inputTokens": 11,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "mps:0/torch.float16": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "driverAllocatedMemoryBytes": 6504267776,
      "memoryMeasurementScope": "Post-inference driver allocation; not peak",
      "torchVersion": "2.10.0",
      "loadMs": 1876.1434170010034,
      "device": "mps",
      "provider": "Metal/MPS",
      "acceleration": "partial",
      "parameterDtype": "torch.float16",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2140.775959000166,
      "grade": "fast",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp16",
    "device": "mps",
    "status": "pass",
    "quality": {
      "expectedPrefix": " x",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 274.46900000359165,
      "inferenceMs": 273.54350000678096,
      "tokensPerSecond": 29.245805511012637,
      "inputTokens": 16,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "mps:0/torch.float16": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "driverAllocatedMemoryBytes": 6512754688,
      "memoryMeasurementScope": "Post-inference driver allocation; not peak",
      "torchVersion": "2.10.0",
      "loadMs": 1876.1434170010034,
      "device": "mps",
      "provider": "Metal/MPS",
      "acceleration": "partial",
      "parameterDtype": "torch.float16",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2150.612417004595,
      "grade": "fast",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " France",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 1152.7111670002341,
      "inferenceMs": 1152.498082999955,
      "tokensPerSecond": 6.941443216266341,
      "inputTokens": 8,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1477.8738749955664,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2630.5850419958006,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " Celsius",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 1199.8001669999212,
      "inferenceMs": 1199.6416249967297,
      "tokensPerSecond": 6.668658233680254,
      "inputTokens": 11,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1477.8738749955664,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2677.6740419954876,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " x",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 1383.1105830031447,
      "inferenceMs": 1382.92429199646,
      "tokensPerSecond": 5.78484306501753,
      "inputTokens": 16,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1477.8738749955664,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2860.984457998711,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " France",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 1329.5288329973118,
      "inferenceMs": 1329.2707910004538,
      "tokensPerSecond": 6.018337312579427,
      "inputTokens": 8,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1430.105667001044,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2759.6344999983557,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " Celsius",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 1184.4219999984489,
      "inferenceMs": 1184.263834002195,
      "tokensPerSecond": 6.755251465346339,
      "inputTokens": 11,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1430.105667001044,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2614.527666999493,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " x",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 1251.8304999975953,
      "inferenceMs": 1251.6379160006181,
      "tokensPerSecond": 6.391624844318034,
      "inputTokens": 16,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1430.105667001044,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2681.9361669986392,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " France",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 1167.1460830038995,
      "inferenceMs": 1166.93933300121,
      "tokensPerSecond": 6.855540621314978,
      "inputTokens": 8,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1506.6771249985322,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2673.823208002432,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " Celsius",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 1187.807290996716,
      "inferenceMs": 1187.57020799967,
      "tokensPerSecond": 6.736443829687435,
      "inputTokens": 11,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1506.6771249985322,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2694.484415995248,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " x",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 4,
      "durationMs": 1259.3396669981303,
      "inferenceMs": 1259.167249998427,
      "tokensPerSecond": 6.353405395518343,
      "inputTokens": 16,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1506.6771249985322,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2766.0167919966625,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " France",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 1,
      "durationMs": 1149.2468340002233,
      "inferenceMs": 1149.0534170006867,
      "tokensPerSecond": 6.962252478115401,
      "inputTokens": 8,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1622.9159170034109,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2772.162751003634,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " Celsius",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 1,
      "durationMs": 1148.1494999970892,
      "inferenceMs": 1148.0004170007305,
      "tokensPerSecond": 6.968638583686951,
      "inputTokens": 11,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1622.9159170034109,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2771.0654170005,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  },
  {
    "precision": "fp32",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "expectedPrefix": " x",
      "prefixPassed": true,
      "streamParity": true,
      "finalParity": true,
      "repetitionPassed": true,
      "passed": true
    },
    "runtime": {
      "engine": "transformers",
      "version": "5.22.1",
      "artifactRevision": "d78a42f79198603e614095753484a04c10c2b940",
      "cpuThreads": 1,
      "durationMs": 1141.495625000971,
      "inferenceMs": 1141.3215420034248,
      "tokensPerSecond": 7.009418210013943,
      "inputTokens": 16,
      "outputTokens": 8,
      "maxNewTokens": 8,
      "placement": {
        "cpu/torch.float32": 326
      },
      "decoding": "greedy",
      "enableThinking": false,
      "outputLimitReached": true,
      "torchVersion": "2.10.0",
      "loadMs": 1622.9159170034109,
      "device": "cpu",
      "provider": "PyTorch CPU",
      "acceleration": "cpu",
      "parameterDtype": "torch.float32",
      "accelerationDetails": "All model parameters and input tensors on selected device; tokenization/decoding on CPU. MPS operator CPU fallback disabled.",
      "totalMs": 2764.4115420043818,
      "grade": "acceptable",
      "loadTimingScope": "Load from pinned local snapshot. Network transfer and runtime provisioning excluded."
    }
  }
]
Release validation still required
  • Public backup and source-outage recovery pending.

Something not quite right?

Tell us what happened with SmolLM3 Desktop. Loading, results, or the page itself—we’re here for all of it.

Attachments Optional

Up to 3 files · 50 MB each

Browser, OS, CPU/GPU, memory, screen, storage, and network details are included automatically. Model inputs, results, and browsing history are not collected.