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.

F2LLM 2

Compare meaning, explore text vectors, and build semantic search on your device.

Browser inferencePowered by WebAI · Private · Unlimited runs
About this model

Convert multilingual text into 640-number vectors for meaning-based retrieval and comparison. Use the vectors to find similar passages; this model does not generate replies.

Created by
CodeFuse
License
Apache-2.0
Model released
Unknown
Source repository created
Mar 9, 2026
Model weights · FP32
637.1 MB

Best for

  • Multilingual semantic search and retrieval.
  • Clustering, similarity, and classification prototypes.

Keep in mind

  • Retrieval quality depends on language, domain, input quality, and correct query instructions.
  • Semantic similarity is not factual verification and can encode training-data bias.
  • Long inputs are truncated and should be chunked.
  • Quantization can change ranking margins.

Processed on your deviceInference runs in this browser. Input/output sharing requires consent for each run. Basic run logs include device information and approximate connection location. Model and runtime files download from external hosts. Saved downloads may be cleared by your browser.

Sources & technical details
Original model
codefuse-ai/F2LLM-v2-160M
Release verification
Unknown
Release evidence
Exact original checkpoint public release date has not been verified. Repository creation, training, paper, artifact conversion and update dates are not release evidence.
Source updated
Sep 3, 2026
Browser model files
onnx-community/F2LLM-v2-160M-ONNX
Files updated
Aug 27, 2026
Artifact revision
7202b3cd72a7f11b4bea90d6c41b67423281360b

Input

TEXT → VECTORS

Sample result

F2LLM-v2 160M — Multilingual Embeddings
Advanced settings 1
Scales each embedding vector to unit length. On makes vector lengths comparable for cosine similarity. Off preserves raw magnitudes only if supported; the listed fixed value must be retained.

Output

Explore what your text means

Generate vectors from one or more texts. Add multiple texts to compare their meaning.

Benchmark evidence

Saved worker evidence is separate from benchmarks run on your device. It does not guarantee speed or quality on every browser.

F2LLM-v2 160M — Multilingual Embeddings

10 configurations tested
Passed4
Failed6
TestedSep 9, 2026
RuntimeTransformers.js 4.2.0

Quality criteria

Finite normalized 640d vectors, distinct rows, and positive cosine exceeds negative by at least 0.1.

PrecisionDeviceResult
FP32GPU · WebGPUAcceleratedPassedFast
Q8GPU · WebGPUAcceleratedPassedFast
FP32CPU · WASMPassedFast
Q8CPU · WASMPassedFast
FP16CPU · WASMFailed—
FP16GPU · WebGPUFailed—
Q4CPU · WASMFailed—
Q4GPU · WebGPUFailed—
Q4F16CPU · WASMFailed—
Q4F16GPU · WebGPUFailed—
Test setup & sample details
Quality gate
Finite normalized 640d vectors, distinct rows, and positive cosine exceeds negative by at least 0.1.
Tested
2026-09-09T19:42:54.587Z
Environment
Not recorded
Runtime
Transformers.js 4.2.0
Artifact revision
7202b3cd72a7f11b4bea90d6c41b67423281360b
Fixture
{
  "file": "f2llm-v2-160m.json",
  "url": "https://f005.backblazeb2.com/file/hubters-webai-public/assets/benchmarks/f2llm-v2-160m.json",
  "size": 576,
  "sha1": "e8a6ac52dcbb23157d108932572d2c44e173e48a"
}
Rerun when
artifact revision change; runtime version change; fixture or quality gate change; browser/backend change
Per-case quality & failure details
[
  {
    "precision": "fp32",
    "device": "wasm",
    "status": "pass",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": 195503.69999992847,
      "inferenceMs": 232.70000004768372,
      "totalMs": 197082.29999995232,
      "embeddingsPerSecond": 12.89213579452194,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "webgpu",
    "status": "pass",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": 209231.5,
      "inferenceMs": 162.79999995231628,
      "totalMs": 210713.10000002384,
      "embeddingsPerSecond": 18.427518432915804,
      "grade": "good"
    }
  },
  {
    "precision": "fp16",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": "Not embedded",
    "failure": "ONNX float16 output type mismatch"
  },
  {
    "precision": "fp16",
    "device": "webgpu",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": "Not embedded",
    "failure": "bounded browser timeout"
  },
  {
    "precision": "q8",
    "device": "wasm",
    "status": "pass",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": 22567,
      "inferenceMs": 245.80000007152557,
      "totalMs": 24192.5,
      "embeddingsPerSecond": 12.2050447482792,
      "grade": "good"
    }
  },
  {
    "precision": "q8",
    "device": "webgpu",
    "status": "pass",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": 22622.300000071526,
      "inferenceMs": 334.5,
      "totalMs": 24250.600000023842,
      "embeddingsPerSecond": 8.968609865470851,
      "grade": "slow"
    }
  },
  {
    "precision": "q4",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": "Not embedded",
    "failure": "bounded browser timeout"
  },
  {
    "precision": "q4",
    "device": "webgpu",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": "Not embedded",
    "failure": "bounded browser timeout"
  },
  {
    "precision": "q4f16",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": "Not embedded",
    "failure": "ONNX float16 output type mismatch"
  },
  {
    "precision": "q4f16",
    "device": "webgpu",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": "Not embedded",
    "failure": "ONNX float16 output type mismatch"
  }
]

F2LLM-v2 80M — Multilingual Embeddings

12 configurations tested
Passed6
Failed6
TestedNov 6, 2026
RuntimeTransformers.js 4.2.0

Quality criteria

Three frozen retrieval fixtures, margins 0.1/0.1/0.05; exact shape, finite unit norms, source/padding parity; each batch <=3s. See frozen PLAN.md.

PrecisionDeviceResult
FP32GPU · WebGPUAcceleratedPassedVery fast
FP32GPU · WebGPUAcceleratedPassedVery fast
FP32GPU · WebGPUAcceleratedPassedFast
FP32CPU · WASMPassedFast
FP32CPU · WASMPassedFast
FP32CPU · WASMPassedFast
INT8GPU · WebGPUFailed—
INT8GPU · WebGPUFailed—
INT8GPU · WebGPUFailed—
INT8CPU · WASMFailed—
INT8CPU · WASMFailed—
INT8CPU · WASMFailed—
Test setup & sample details
Quality gate
Three frozen retrieval fixtures, margins 0.1/0.1/0.05; exact shape, finite unit norms, source/padding parity; each batch <=3s. See frozen PLAN.md.
Tested
2026-11-06T12:01:05.922Z
Environment
Not recorded
Runtime
Transformers.js 4.2.0
Artifact revision
6dcc8939ec423ce1e8b9d159b633aae63e86378d
Fixture
{
  "file": "model-samples/f2llm-v2-english.json",
  "size": 676,
  "sha1": "19092a40118140df3189183de0231b3b39636667",
  "sha256": "841050ddaa821a11c33413cb97d7498c4010165fbe9d3f1247e32b009ba7296f",
  "url": "https://assets.localai.tools/samples/v1/files/841050ddaa821a11c33413cb97d7498c4010165fbe9d3f1247e32b009ba7296f/f2llm-v2-english.json"
}
Rerun when
artifact revision change; runtime version change; fixture or quality gate change; browser/backend change
Per-case quality & failure details
[
  {
    "precision": "fp32",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "valid": true,
      "margin": 0.7883882574173738,
      "minimumMargin": 0.1,
      "norms": [
        0.9999998619700385,
        1.0000002935109795,
        0.9999999389743709
      ],
      "sourceCos": [
        0.9999999999987704,
        0.9999999999992553,
        0.999999999998423
      ],
      "sourceDelta": [
        3.241002559661865e-7,
        3.129243850708008e-7,
        3.2782554626464844e-7
      ],
      "batchCos": [
        1,
        1,
        1
      ],
      "batchDelta": [
        0,
        0,
        0
      ],
      "rawCos": [
        1.0000000000000004,
        0.9999999999999992,
        1.0000000000000004
      ],
      "passed": true
    },
    "runtime": {
      "loadMs": 4129,
      "inferenceMs": 123.89999999990687,
      "textsPerSecond": 24.21307506055089,
      "grade": "good",
      "embeddingsPerSecond": 24.21307506055089,
      "totalMs": 5844
    }
  },
  {
    "precision": "fp32",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "valid": true,
      "margin": 0.5921693095642787,
      "minimumMargin": 0.1,
      "norms": [
        1.000000217638959,
        0.999999979019631,
        1.0000001551768276
      ],
      "sourceCos": [
        0.999999999998882,
        0.9999999999991482,
        0.9999999999986395
      ],
      "sourceDelta": [
        2.682209014892578e-7,
        2.384185791015625e-7,
        3.3155083656311035e-7
      ],
      "batchCos": [
        1,
        1,
        1
      ],
      "batchDelta": [
        0,
        0,
        0
      ],
      "rawCos": [
        1.0000000000000009,
        1.0000000000000009,
        1.0000000000000016
      ],
      "passed": true
    },
    "runtime": {
      "loadMs": 4129,
      "inferenceMs": 22.800000000279397,
      "textsPerSecond": 131.57894736680865,
      "grade": "excellent",
      "embeddingsPerSecond": 131.57894736680865,
      "totalMs": 5844
    }
  },
  {
    "precision": "fp32",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "valid": true,
      "margin": 0.05289184862134544,
      "minimumMargin": 0.05,
      "norms": [
        0.9999999505679673,
        0.9999999784280017,
        1.0000001670593943
      ],
      "sourceCos": [
        0.9999999999987494,
        0.9999999999989397,
        0.9999999999991143
      ],
      "sourceDelta": [
        2.8312206268310547e-7,
        2.682209014892578e-7,
        2.384185791015625e-7
      ],
      "batchCos": [
        1,
        0.9999999999994145,
        0.9999999999993898
      ],
      "batchDelta": [
        0,
        2.2351741790771484e-7,
        2.5331974029541016e-7
      ],
      "rawCos": [
        1.000000000000002,
        1.0000000000000002,
        1.0000000000000004
      ],
      "passed": true
    },
    "runtime": {
      "loadMs": 4129,
      "inferenceMs": 20.899999999906868,
      "textsPerSecond": 143.54066985709895,
      "grade": "excellent",
      "embeddingsPerSecond": 143.54066985709895,
      "totalMs": 5844
    }
  },
  {
    "precision": "fp32",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "valid": true,
      "margin": 0.7883882953360566,
      "minimumMargin": 0.1,
      "norms": [
        0.9999996670392678,
        1.000000144397986,
        0.9999999843129589
      ],
      "sourceCos": [
        0.9999999999978856,
        0.9999999999988345,
        0.9999999999984658
      ],
      "sourceDelta": [
        4.6193599700927734e-7,
        2.980232238769531e-7,
        3.296881914138794e-7
      ],
      "batchCos": [
        1,
        0.9999999999998772,
        0.999999999999895
      ],
      "batchDelta": [
        0,
        8.568167686462402e-8,
        8.66129994392395e-8
      ],
      "rawCos": [
        1.0000000000000002,
        1.0000000000000013,
        1.0000000000000018
      ],
      "passed": true
    },
    "runtime": {
      "loadMs": 3174.5,
      "inferenceMs": 139.5,
      "textsPerSecond": 21.50537634408602,
      "grade": "good",
      "embeddingsPerSecond": 21.50537634408602,
      "totalMs": 5806.600000000093
    }
  },
  {
    "precision": "fp32",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "valid": true,
      "margin": 0.5921692782017401,
      "minimumMargin": 0.1,
      "norms": [
        1.0000000910763205,
        0.9999998437896036,
        1.0000000459740335
      ],
      "sourceCos": [
        0.9999999999986027,
        0.9999999999987887,
        0.9999999999985216
      ],
      "sourceDelta": [
        4.023313522338867e-7,
        2.980232238769531e-7,
        3.129243850708008e-7
      ],
      "batchCos": [
        1,
        1,
        0.9999999999998974
      ],
      "batchDelta": [
        0,
        0,
        8.940696716308594e-8
      ],
      "rawCos": [
        1.000000000000002,
        1.0000000000000002,
        1
      ],
      "passed": true
    },
    "runtime": {
      "loadMs": 3174.5,
      "inferenceMs": 111.3000000002794,
      "textsPerSecond": 26.95417789750646,
      "grade": "good",
      "embeddingsPerSecond": 26.95417789750646,
      "totalMs": 5806.600000000093
    }
  },
  {
    "precision": "fp32",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "valid": true,
      "margin": 0.05289174754590553,
      "minimumMargin": 0.05,
      "norms": [
        0.999999911549801,
        1.000000029297134,
        1.0000000049615525
      ],
      "sourceCos": [
        0.999999999998139,
        0.999999999999042,
        0.9999999999988328
      ],
      "sourceDelta": [
        3.948807716369629e-7,
        2.871965989470482e-7,
        4.172325134277344e-7
      ],
      "batchCos": [
        1,
        0.9999999999999039,
        0.9999999999999083
      ],
      "batchDelta": [
        0,
        7.82310962677002e-8,
        7.82310962677002e-8
      ],
      "rawCos": [
        0.9999999999999996,
        1.0000000000000002,
        1.0000000000000009
      ],
      "passed": true
    },
    "runtime": {
      "loadMs": 3174.5,
      "inferenceMs": 161.60000000009313,
      "textsPerSecond": 18.564356435632867,
      "grade": "good",
      "embeddingsPerSecond": 18.564356435632867,
      "totalMs": 5806.600000000093
    }
  },
  {
    "precision": "int8",
    "device": "webgpu",
    "status": "fail",
    "quality": {
      "valid": true,
      "margin": 0.7651009386642273,
      "minimumMargin": 0.1,
      "norms": [
        1.0000000414169783,
        1.0000002098540641,
        0.9999999796575847
      ],
      "sourceCos": [
        0.9876063346562519,
        0.9901754916114575,
        0.9893153316476975
      ],
      "sourceDelta": [
        0.025913090212270617,
        0.03127509169280529,
        0.02465012203902006
      ],
      "batchCos": [
        0.987198691490626,
        0.9908676495512938,
        0.9892802310892941
      ],
      "batchDelta": [
        0.034569825045764446,
        0.030992217594757676,
        0.025083132088184357
      ],
      "rawCos": [
        1,
        1.0000000000000018,
        1.0000000000000004
      ],
      "passed": false
    },
    "runtime": {
      "loadMs": 2318.899999999907,
      "inferenceMs": 239.89999999990687,
      "textsPerSecond": 12.505210504381678,
      "grade": "good",
      "embeddingsPerSecond": 12.505210504381678,
      "totalMs": 5676.700000000186
    }
  },
  {
    "precision": "int8",
    "device": "webgpu",
    "status": "fail",
    "quality": {
      "valid": true,
      "margin": 0.5429080281767857,
      "minimumMargin": 0.1,
      "norms": [
        0.9999999517584289,
        0.999999907991937,
        0.9999998276630089
      ],
      "sourceCos": [
        0.9868189381130221,
        0.9882312273568769,
        0.9892193826324432
      ],
      "sourceDelta": [
        0.027224838733673096,
        0.026743966154754162,
        0.030383698642253876
      ],
      "batchCos": [
        0.9831069120798801,
        0.989229327923064,
        0.9891307958090879
      ],
      "batchDelta": [
        0.03839692287147045,
        0.029526042751967907,
        0.025980941951274872
      ],
      "rawCos": [
        1.0000000000000009,
        1.000000000000001,
        1.000000000000001
      ],
      "passed": false
    },
    "runtime": {
      "loadMs": 2318.899999999907,
      "inferenceMs": 159.60000000009313,
      "textsPerSecond": 18.79699248119204,
      "grade": "good",
      "embeddingsPerSecond": 18.79699248119204,
      "totalMs": 5676.700000000186
    }
  },
  {
    "precision": "int8",
    "device": "webgpu",
    "status": "fail",
    "quality": {
      "valid": true,
      "margin": 0.05507841055333007,
      "minimumMargin": 0.05,
      "norms": [
        1.0000000134435432,
        1.0000000970872212,
        1.0000002818021494
      ],
      "sourceCos": [
        0.9807520338869736,
        0.9902007532318311,
        0.9854899392277647
      ],
      "sourceDelta": [
        0.02882520528510213,
        0.02262839861214161,
        0.03710826486349106
      ],
      "batchCos": [
        0.9825097696200873,
        0.9909646187255378,
        0.9879942931497481
      ],
      "batchDelta": [
        0.03345338674262166,
        0.029254399240016937,
        0.027618792781140655
      ],
      "rawCos": [
        0.9999999999999993,
        1,
        1.0000000000000007
      ],
      "passed": false
    },
    "runtime": {
      "loadMs": 2318.899999999907,
      "inferenceMs": 213,
      "textsPerSecond": 14.084507042253522,
      "grade": "good",
      "embeddingsPerSecond": 14.084507042253522,
      "totalMs": 5676.700000000186
    }
  },
  {
    "precision": "int8",
    "device": "wasm",
    "status": "fail",
    "quality": {
      "valid": true,
      "margin": 0.7946359962127176,
      "minimumMargin": 0.1,
      "norms": [
        1.0000000827749644,
        1.0000000895825332,
        1.0000001143306254
      ],
      "sourceCos": [
        0.9867364964639963,
        0.9904088049055554,
        0.9887428829110436
      ],
      "sourceDelta": [
        0.03211348503828049,
        0.02534858137369156,
        0.025501442898530513
      ],
      "batchCos": [
        0.9883233284058318,
        0.9912322617783567,
        0.9892581098424044
      ],
      "batchDelta": [
        0.047146543860435486,
        0.021313898265361786,
        0.027540819719433784
      ],
      "rawCos": [
        1.0000000000000007,
        0.9999999999999992,
        1.000000000000001
      ],
      "passed": false
    },
    "runtime": {
      "loadMs": 2284.600000000093,
      "inferenceMs": 146.39999999990687,
      "textsPerSecond": 20.49180327870156,
      "grade": "good",
      "embeddingsPerSecond": 20.49180327870156,
      "totalMs": 4902.700000000186
    }
  },
  {
    "precision": "int8",
    "device": "wasm",
    "status": "fail",
    "quality": {
      "valid": true,
      "margin": 0.5578629745941899,
      "minimumMargin": 0.1,
      "norms": [
        0.9999998461681031,
        1.0000002098675147,
        1.0000003432205142
      ],
      "sourceCos": [
        0.9864657111246161,
        0.9856634726630936,
        0.9898322657582609
      ],
      "sourceDelta": [
        0.03131098486483097,
        0.03658842481672764,
        0.026654399931430817
      ],
      "batchCos": [
        0.9820268110064048,
        0.9905007682209811,
        0.9890373185878334
      ],
      "batchDelta": [
        0.04161050729453564,
        0.02442561835050583,
        0.024205350317060947
      ],
      "rawCos": [
        1.0000000000000002,
        0.9999999999999998,
        1.0000000000000009
      ],
      "passed": false
    },
    "runtime": {
      "loadMs": 2284.600000000093,
      "inferenceMs": 115.5,
      "textsPerSecond": 25.974025974025974,
      "grade": "good",
      "embeddingsPerSecond": 25.974025974025974,
      "totalMs": 4902.700000000186
    }
  },
  {
    "precision": "int8",
    "device": "wasm",
    "status": "fail",
    "quality": {
      "valid": true,
      "margin": 0.0620554974762173,
      "minimumMargin": 0.05,
      "norms": [
        1.0000000357621823,
        0.9999995853632814,
        1.0000000585394115
      ],
      "sourceCos": [
        0.9853424163505962,
        0.9885563146088948,
        0.9878784987580828
      ],
      "sourceDelta": [
        0.029380179941654205,
        0.023128945380449295,
        0.02605897723697126
      ],
      "batchCos": [
        0.9838618014219879,
        0.9894261012360178,
        0.9897605666246495
      ],
      "batchDelta": [
        0.02834775298833847,
        0.023736491799354553,
        0.024677593261003494
      ],
      "rawCos": [
        1.0000000000000007,
        1.0000000000000002,
        1.0000000000000009
      ],
      "passed": false
    },
    "runtime": {
      "loadMs": 2284.600000000093,
      "inferenceMs": 166.79999999981374,
      "textsPerSecond": 17.985611510811452,
      "grade": "good",
      "embeddingsPerSecond": 17.985611510811452,
      "totalMs": 4902.700000000186
    }
  }
]

Something not quite right?

Tell us what happened with F2LLM 2. 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.