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.

Liquid LFM3 Desktop

Local text chat with Liquid LFM3 1.7B.

Desktop inferencePrivate · Runs on your computer
About this model

Answer typed classification questions about text or one media attachment.

Created by
Liquid AI
Model released
Unknown
Source repository created
Oct 24, 2026
Model weights · Q4
1.3 GB

Best for

  • Choose a label, assign an ordered score, or answer a yes/no probability question.

Keep in mind

  • Verify classifications against task-specific references. Confidence and probabilities may be poorly calibrated.
  • The allowed media input is one PNG.
  • Video and sampled frames are excluded by the product policy.
  • Commercial use and redistribution are subject to the Liquid AI license.

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

Q4 · metal (verified platform only)

Sources & technical details
Original model
https://huggingface.co/LiquidAI/d1-2B
Release verification
Unknown
Release evidence
Exact checkpoint date pending provenance review.
Source updated
Nov 14, 2026
Native model files
LiquidAI/d1-2B-GGUF
Files updated
Nov 4, 2026
Artifact revision
4a181094d1b6addfaa3a0956f78e7ab2aa406a0c

Desktop app

Sample result

d1-2B Desktop

Run Liquid LFM3 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 / Q4TestedNot establishedRAM estimate
macOS Apple SiliconMETAL / Q4TestedNot establishedRAM estimate
If your device falls short

MemoryShow advisory requirements.

GPUUse verified CPU recovery when permitted.

PlatformNo verified configuration.

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.

d1-2B Desktop

6 configurations tested
Passed6
Failed0
TestedNov 15, 2026
Runtimellama.cpp b12359

Quality criteria

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

PrecisionDeviceResultPlatformInference / load
Q4metalAcceleratedPassedFastmacos · aarch64134 ms / 869 ms
Q4metalAcceleratedPassedFastmacos · aarch64137 ms / 869 ms
Q4metalAcceleratedPassedFastmacos · aarch64140 ms / 869 ms
Q4cpuPassedFastmacos · aarch64157 ms / 7097 ms
Q4cpuPassedFastmacos · aarch64450 ms / 7097 ms
Q4cpuPassedFastmacos · aarch64541 ms / 7097 ms
Test setup & sample details
Quality gate
No explicit quality gate recorded. Review the per-case quality evidence; full certification is pending.
Tested
2026-11-15T22:45:38.686Z
Environment
{ "scope": "DesktopAI TypeScript client using native JSON-lines transport with verified cached sources.", "hardware": "Apple M3 Max", "os": "macOS", "arch": "arm64", "runtime": { "name": "llama.cpp", "version": "b12359" }, "speedGradePolicy": "Short requests completing in 2 seconds or less are interactive." }
Runtime
llama.cpp b12359
Artifact revision
4a181094d1b6addfaa3a0956f78e7ab2aa406a0c
Fixture
{
  "file": "model-samples/lfm3-d1-2b-native/d1-native-samples/text-choice-request.json",
  "url": "https://assets.localai.tools/samples/v1/files/4c95fd8e06b5a0b2695aa67c27cae0c49060eda43f7f474c9f0a597c654c3a28/text-choice-request.json",
  "size": 382,
  "sha1": "43475bf5ce0e9e77d8c8b6b4bf80bf99361b7952",
  "sha256": "4c95fd8e06b5a0b2695aa67c27cae0c49060eda43f7f474c9f0a597c654c3a28"
}
Rerun when
artifactRevision; runtimeVersion; fixtureChecksum; qualityGate; nativeBackend
Per-case quality & failure details
[
  {
    "precision": "q4",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "choice": "Billing"
        }
      ]
    },
    "runtime": {
      "acceleration": "cpu",
      "cpuThreads": 4,
      "device": "cpu",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 156.5475,
      "loadMs": 7097.4473339999995,
      "offloadedLayers": 0,
      "precision": "q4",
      "provider": "llama.cpp CPU",
      "requestedDevice": "cpu",
      "requestedPrecision": "q4",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "totalMs": 7253.994833999999,
      "artifactRevision": "4a181094d1b6addfaa3a0956f78e7ab2aa406a0c",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 6.387837557290918
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q4",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "choice": "Red"
        }
      ]
    },
    "runtime": {
      "acceleration": "cpu",
      "cpuThreads": 4,
      "device": "cpu",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 541.386583,
      "loadMs": 7097.4473339999995,
      "offloadedLayers": 0,
      "precision": "q4",
      "provider": "llama.cpp CPU",
      "requestedDevice": "cpu",
      "requestedPrecision": "q4",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "visualizationMs": 0.021542000000000002,
      "totalMs": 7638.833917,
      "artifactRevision": "4a181094d1b6addfaa3a0956f78e7ab2aa406a0c",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 1.8471089447002422
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q4",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "noul": [
            0,
            0.2
          ]
        },
        {
          "score": [
            2.5,
            3
          ]
        },
        {
          "noul": [
            0.8,
            1
          ]
        }
      ]
    },
    "runtime": {
      "acceleration": "cpu",
      "cpuThreads": 4,
      "device": "cpu",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 450.436417,
      "loadMs": 7097.4473339999995,
      "offloadedLayers": 0,
      "precision": "q4",
      "provider": "llama.cpp CPU",
      "requestedDevice": "cpu",
      "requestedPrecision": "q4",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "totalMs": 7547.883750999999,
      "artifactRevision": "4a181094d1b6addfaa3a0956f78e7ab2aa406a0c",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 6.660207493835917
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q4",
    "device": "metal",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "choice": "Billing"
        }
      ]
    },
    "runtime": {
      "acceleration": "partial",
      "cpuThreads": 4,
      "device": "metal",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 136.561875,
      "loadMs": 868.922,
      "offloadedLayers": 25,
      "precision": "q4",
      "provider": "llama.cpp Metal",
      "requestedDevice": "metal",
      "requestedPrecision": "q4",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "totalMs": 1005.483875,
      "artifactRevision": "4a181094d1b6addfaa3a0956f78e7ab2aa406a0c",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 7.322687975688677
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q4",
    "device": "metal",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "choice": "Red"
        }
      ]
    },
    "runtime": {
      "acceleration": "partial",
      "cpuThreads": 4,
      "device": "metal",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 133.643667,
      "loadMs": 868.922,
      "offloadedLayers": 25,
      "precision": "q4",
      "provider": "llama.cpp Metal",
      "requestedDevice": "metal",
      "requestedPrecision": "q4",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "visualizationMs": 0.037709,
      "totalMs": 1002.5656670000001,
      "artifactRevision": "4a181094d1b6addfaa3a0956f78e7ab2aa406a0c",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 7.482584266413462
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q4",
    "device": "metal",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "noul": [
            0,
            0.2
          ]
        },
        {
          "score": [
            2.5,
            3
          ]
        },
        {
          "noul": [
            0.8,
            1
          ]
        }
      ]
    },
    "runtime": {
      "acceleration": "partial",
      "cpuThreads": 4,
      "device": "metal",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 139.911,
      "loadMs": 868.922,
      "offloadedLayers": 25,
      "precision": "q4",
      "provider": "llama.cpp Metal",
      "requestedDevice": "metal",
      "requestedPrecision": "q4",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "totalMs": 1008.8330000000001,
      "artifactRevision": "4a181094d1b6addfaa3a0956f78e7ab2aa406a0c",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 21.442202543045223
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  }
]

d1-Omni-2B Desktop

8 configurations tested
Passed8
Failed0
TestedNov 15, 2026
Runtimellama.cpp b12359

Quality criteria

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

PrecisionDeviceResultPlatformInference / load
Q8metalAcceleratedPassedFastmacos · aarch64128 ms / 760 ms
Q8metalAcceleratedPassedFastmacos · aarch64161 ms / 760 ms
Q8metalAcceleratedPassedFastmacos · aarch64189 ms / 760 ms
Q8metalAcceleratedPassedFastmacos · aarch64202 ms / 760 ms
Q8cpuPassedFastmacos · aarch64173 ms / 7642 ms
Q8cpuPassedFastmacos · aarch64495 ms / 7642 ms
Q8cpuPassedFastmacos · aarch64545 ms / 7642 ms
Q8cpuPassedFastmacos · aarch64734 ms / 7642 ms
Test setup & sample details
Quality gate
No explicit quality gate recorded. Review the per-case quality evidence; full certification is pending.
Tested
2026-11-15T22:45:51.616Z
Environment
{ "scope": "DesktopAI TypeScript client using native JSON-lines transport with verified cached sources.", "hardware": "Apple M3 Max", "os": "macOS", "arch": "arm64", "runtime": { "name": "llama.cpp", "version": "b12359" }, "speedGradePolicy": "Short requests completing in 2 seconds or less are interactive." }
Runtime
llama.cpp b12359
Artifact revision
36707287af55fc9ab03e34bda6a6cc82075499b1
Fixture
{
  "file": "model-samples/lfm3-d1-omni-2b-native/d1-omni-native-samples/text-choice-request.json",
  "url": "https://assets.localai.tools/samples/v1/files/4c95fd8e06b5a0b2695aa67c27cae0c49060eda43f7f474c9f0a597c654c3a28/text-choice-request.json",
  "size": 382,
  "sha1": "43475bf5ce0e9e77d8c8b6b4bf80bf99361b7952",
  "sha256": "4c95fd8e06b5a0b2695aa67c27cae0c49060eda43f7f474c9f0a597c654c3a28"
}
Rerun when
artifactRevision; runtimeVersion; fixtureChecksum; qualityGate; nativeBackend
Per-case quality & failure details
[
  {
    "precision": "q8",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "choice": "Billing"
        }
      ]
    },
    "runtime": {
      "acceleration": "cpu",
      "cpuThreads": 4,
      "device": "cpu",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 172.592833,
      "loadMs": 7642.264125,
      "offloadedLayers": 0,
      "precision": "q8",
      "provider": "llama.cpp CPU",
      "requestedDevice": "cpu",
      "requestedPrecision": "q8",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "totalMs": 7814.856957999999,
      "artifactRevision": "36707287af55fc9ab03e34bda6a6cc82075499b1",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 5.793983345762682
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q8",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "choice": "Red"
        }
      ]
    },
    "runtime": {
      "acceleration": "cpu",
      "cpuThreads": 4,
      "device": "cpu",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 544.943541,
      "loadMs": 7642.264125,
      "offloadedLayers": 0,
      "precision": "q8",
      "provider": "llama.cpp CPU",
      "requestedDevice": "cpu",
      "requestedPrecision": "q8",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "visualizationMs": 0.017209000000000002,
      "totalMs": 8187.207665999999,
      "artifactRevision": "36707287af55fc9ab03e34bda6a6cc82075499b1",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 1.8350524866575124
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q8",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "noul": [
            0,
            0.2
          ]
        },
        {
          "score": [
            2.5,
            3
          ]
        },
        {
          "noul": [
            0.8,
            1
          ]
        }
      ]
    },
    "runtime": {
      "acceleration": "cpu",
      "cpuThreads": 4,
      "device": "cpu",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 494.640041,
      "loadMs": 7642.264125,
      "offloadedLayers": 0,
      "precision": "q8",
      "provider": "llama.cpp CPU",
      "requestedDevice": "cpu",
      "requestedPrecision": "q8",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "totalMs": 8136.904165999999,
      "artifactRevision": "36707287af55fc9ab03e34bda6a6cc82075499b1",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 6.065016479327035
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q8",
    "device": "cpu",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "choice": "Americans"
        }
      ]
    },
    "runtime": {
      "acceleration": "cpu",
      "cpuThreads": 4,
      "device": "cpu",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 733.6882919999999,
      "loadMs": 7642.264125,
      "offloadedLayers": 0,
      "precision": "q8",
      "provider": "llama.cpp CPU",
      "requestedDevice": "cpu",
      "requestedPrecision": "q8",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "totalMs": 8375.952417,
      "artifactRevision": "36707287af55fc9ab03e34bda6a6cc82075499b1",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 1.3629766358599602
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q8",
    "device": "metal",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "choice": "Billing"
        }
      ]
    },
    "runtime": {
      "acceleration": "partial",
      "cpuThreads": 4,
      "device": "metal",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 128.40520899999999,
      "loadMs": 759.668208,
      "offloadedLayers": 25,
      "precision": "q8",
      "provider": "llama.cpp Metal",
      "requestedDevice": "metal",
      "requestedPrecision": "q8",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "totalMs": 888.0734170000001,
      "artifactRevision": "36707287af55fc9ab03e34bda6a6cc82075499b1",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 7.787846052257897
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q8",
    "device": "metal",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "choice": "Red"
        }
      ]
    },
    "runtime": {
      "acceleration": "partial",
      "cpuThreads": 4,
      "device": "metal",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 188.50583300000002,
      "loadMs": 759.668208,
      "offloadedLayers": 25,
      "precision": "q8",
      "provider": "llama.cpp Metal",
      "requestedDevice": "metal",
      "requestedPrecision": "q8",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "visualizationMs": 0.024666,
      "totalMs": 948.1740410000001,
      "artifactRevision": "36707287af55fc9ab03e34bda6a6cc82075499b1",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 5.304875632150862
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q8",
    "device": "metal",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "noul": [
            0,
            0.2
          ]
        },
        {
          "score": [
            2.5,
            3
          ]
        },
        {
          "noul": [
            0.8,
            1
          ]
        }
      ]
    },
    "runtime": {
      "acceleration": "partial",
      "cpuThreads": 4,
      "device": "metal",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 201.661083,
      "loadMs": 759.668208,
      "offloadedLayers": 25,
      "precision": "q8",
      "provider": "llama.cpp Metal",
      "requestedDevice": "metal",
      "requestedPrecision": "q8",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "totalMs": 961.329291,
      "artifactRevision": "36707287af55fc9ab03e34bda6a6cc82075499b1",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 14.876444950957643
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  },
  {
    "precision": "q8",
    "device": "metal",
    "status": "pass",
    "quality": {
      "metric": "frozenDecisionGate",
      "score": 1,
      "passed": true,
      "gates": [
        {
          "choice": "Americans"
        }
      ]
    },
    "runtime": {
      "acceleration": "partial",
      "cpuThreads": 4,
      "device": "metal",
      "engine": "llama.cpp",
      "fallback": false,
      "fallbackReason": null,
      "inferenceMs": 160.55625,
      "loadMs": 759.668208,
      "offloadedLayers": 25,
      "precision": "q8",
      "provider": "llama.cpp Metal",
      "requestedDevice": "metal",
      "requestedPrecision": "q8",
      "selectionReason": "Pinned package with verified layer placement",
      "version": "b12359",
      "totalMs": 920.224458,
      "artifactRevision": "36707287af55fc9ab03e34bda6a6cc82075499b1",
      "grade": "good",
      "performanceMetric": {
        "name": "questionsPerSecond",
        "value": 6.228346763206042
      },
      "platform": "macos",
      "arch": "aarch64"
    }
  }
]
Release validation still required
  • Repository-wide production lint retains seven pre-existing generated DuckDB errors; strict Clippy retains seven unrelated adapter errors.

Something not quite right?

Tell us what happened with Liquid LFM3 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.