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.

LateOn Code Edge 17M

Encode natural-language queries and source-code snippets as 48-number vectors for each retained token.

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

Encode natural-language queries and source-code snippets as 48-number vectors for each retained token.

Created by
LightOn AI
License
Apache-2.0
Model released
Feb 12, 2026
Source repository created
Feb 5, 2026
Model weights · FP32
68.0 MB

Best for

  • Late-interaction text retrieval using MaxSim.
  • Token-level query and passage comparison.

Keep in mind

  • Measured on three small code retrieval fixtures. Evaluate on your own corpus before relying on rankings.
  • Queries are truncated to 256 tokens and documents to 2048 tokens, including special tokens.
  • Retrieval scores measure relevance and can reflect training-data biases. Scores do not establish factual accuracy.
  • The selected runtime returns complete token matrices after inference. Incremental streaming is unavailable.

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
lightonai/LateOn-Code-edge
Release verification
Verified
Release evidence
LightOn’s dated release announcement identifies the exact checkpoint and links the pretrained baselines.
Source updated
Aug 17, 2026
Browser model files
lightonai/LateOn-Code-edge
Files updated
Aug 17, 2026
Artifact revision
4bcdf5ed93f791259eb130b577a240f753d68dd8

Input

TEXT

Sample result

LateOn Code Edge 17M

Settings

Choose the encoder prefix and token limit. Uses the matching query or document token policy.

Output

Your result will appear here

Add input or try the sample, then run the model.

Benchmark evidence

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

LateOn Code Edge 17M

2 configurations tested
Passed2
Failed0
TestedNov 14, 2026
RuntimeONNX Runtime Web 1.24.3

Quality criteria

Three authored code-ranking fixtures, exact token IDs and agreement with original native FP32 token vectors on browser.

PrecisionDeviceResult
FP32GPU · WebGPUAcceleratedPassedFast
FP32CPU · WASMPassedVery fast
Test setup & sample details
Quality gate
Three authored code-ranking fixtures, exact token IDs and agreement with original native FP32 token vectors on browser.
Tested
2026-11-14T12:51:24.843Z
Environment
Apple M3 Max on macOS 27.2. Managed native or isolated Chrome module workers.
Runtime
ONNX Runtime Web 1.24.3
Artifact revision
4bcdf5ed93f791259eb130b577a240f753d68dd8
Fixture
{
  "file": "model-samples/lateon-code-deduplicate-query.json",
  "size": 193,
  "sha1": "aa4a1de14636660ffe22aa8f370645d65e11cace",
  "sha256": "92f8a217d38938179b0d9c1f40bab01fbdd3aba3d593558f706a7d4714205e8a",
  "url": "https://assets.localai.tools/samples/v1/files/92f8a217d38938179b0d9c1f40bab01fbdd3aba3d593558f706a7d4714205e8a/lateon-code-deduplicate-query.json",
  "durationSeconds": 0
}
Rerun when
artifact revision; runtime version; retrieval fixture; tokenizer or projection semantics
Per-case quality & failure details
[
  {
    "precision": "fp32",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "passed": true,
      "tokenParity": true,
      "winner": 0,
      "expected": 0,
      "minimumTokenCosine": 0.9999999346137198,
      "scores": [
        8.12368242077584,
        7.640633572152186,
        6.96742971534647
      ],
      "grade": "retrieval-pass"
    },
    "runtime": {
      "durationMs": 34.299999952316284,
      "inferenceMs": 34.299999952316284,
      "loadMs": 1820.199999988079,
      "totalMs": 1854.4999999403954,
      "embeddingsPerSecond": 29.15451899096781,
      "grade": "interactive",
      "engine": "ONNX Runtime Web",
      "version": "1.24.3",
      "device": "wasm",
      "precision": "fp32",
      "fallback": false,
      "artifactRevision": "4bcdf5ed93f791259eb130b577a240f753d68dd8",
      "acceleration": "cpu"
    }
  },
  {
    "precision": "fp32",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "tokenParity": true,
      "winner": 0,
      "expected": 0,
      "minimumTokenCosine": 0.9999999346138353,
      "scores": [
        8.123683150005103,
        7.640632578055829,
        6.967429626552582
      ],
      "grade": "retrieval-pass"
    },
    "runtime": {
      "durationMs": 116.90000003576279,
      "inferenceMs": 116.90000003576279,
      "loadMs": 1567.7000000476837,
      "totalMs": 1684.6000000834465,
      "embeddingsPerSecond": 8.554319928948448,
      "grade": "interactive",
      "engine": "ONNX Runtime Web",
      "version": "1.24.3",
      "device": "webgpu",
      "precision": "fp32",
      "fallback": false,
      "artifactRevision": "4bcdf5ed93f791259eb130b577a240f753d68dd8",
      "acceleration": "partial",
      "gpuSubmissionObserved": true,
      "accelerationDetails": "WebGPU graph execution observed. Full graph GPU placement was not measured."
    }
  }
]

LateOn Code Base

4 configurations tested
Passed4
Failed0
TestedNov 14, 2026
RuntimeONNX Runtime Web 1.24.3

Quality criteria

Three authored code-ranking fixtures, exact token IDs and agreement with original native FP32 token vectors on browser.

PrecisionDeviceResult
FP32GPU · WebGPUAcceleratedPassedFast
INT8GPU · WebGPUAcceleratedPassedFast
FP32CPU · WASMPassedFast
INT8CPU · WASMPassedFast
Test setup & sample details
Quality gate
Three authored code-ranking fixtures, exact token IDs and agreement with original native FP32 token vectors on browser.
Tested
2026-11-14T12:51:42.802Z
Environment
Apple M3 Max on macOS 27.2. Managed native or isolated Chrome module workers.
Runtime
ONNX Runtime Web 1.24.3
Artifact revision
ace431824f35db231178fc602e33296784762a2e
Fixture
{
  "file": "model-samples/lateon-code-deduplicate-query.json",
  "size": 193,
  "sha1": "aa4a1de14636660ffe22aa8f370645d65e11cace",
  "sha256": "92f8a217d38938179b0d9c1f40bab01fbdd3aba3d593558f706a7d4714205e8a",
  "url": "https://assets.localai.tools/samples/v1/files/92f8a217d38938179b0d9c1f40bab01fbdd3aba3d593558f706a7d4714205e8a/lateon-code-deduplicate-query.json",
  "durationSeconds": 0
}
Rerun when
artifact revision; runtime version; retrieval fixture; tokenizer or projection semantics
Per-case quality & failure details
[
  {
    "precision": "int8",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "passed": true,
      "tokenParity": true,
      "winner": 0,
      "expected": 0,
      "minimumTokenCosine": 0.9785590116670111,
      "scores": [
        7.128134502669535,
        6.460837637835252,
        4.538000073936346
      ],
      "grade": "retrieval-pass"
    },
    "runtime": {
      "durationMs": 116.79999995231628,
      "inferenceMs": 116.79999995231628,
      "loadMs": 2065,
      "totalMs": 2181.7999999523163,
      "embeddingsPerSecond": 8.561643839111738,
      "grade": "interactive",
      "engine": "ONNX Runtime Web",
      "version": "1.24.3",
      "device": "wasm",
      "precision": "int8",
      "fallback": false,
      "artifactRevision": "ace431824f35db231178fc602e33296784762a2e",
      "acceleration": "cpu"
    }
  },
  {
    "precision": "int8",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "tokenParity": true,
      "winner": 0,
      "expected": 0,
      "minimumTokenCosine": 0.9684123226630045,
      "scores": [
        7.190391574343532,
        6.627473275337507,
        4.564985630364577
      ],
      "grade": "retrieval-pass"
    },
    "runtime": {
      "durationMs": 299.7999999523163,
      "inferenceMs": 299.7999999523163,
      "loadMs": 2282.100000023842,
      "totalMs": 2581.899999976158,
      "embeddingsPerSecond": 3.3355570385558764,
      "grade": "slow",
      "engine": "ONNX Runtime Web",
      "version": "1.24.3",
      "device": "webgpu",
      "precision": "int8",
      "fallback": false,
      "artifactRevision": "ace431824f35db231178fc602e33296784762a2e",
      "acceleration": "partial",
      "gpuSubmissionObserved": true,
      "accelerationDetails": "WebGPU graph execution observed. Full graph GPU placement was not measured."
    }
  },
  {
    "precision": "fp32",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "passed": true,
      "tokenParity": true,
      "winner": 0,
      "expected": 0,
      "minimumTokenCosine": 0.9999999148196693,
      "scores": [
        7.260654897623601,
        6.410420206507405,
        4.511354022720276
      ],
      "grade": "retrieval-pass"
    },
    "runtime": {
      "durationMs": 104,
      "inferenceMs": 104,
      "loadMs": 5061.200000047684,
      "totalMs": 5165.200000047684,
      "embeddingsPerSecond": 9.615384615384615,
      "grade": "interactive",
      "engine": "ONNX Runtime Web",
      "version": "1.24.3",
      "device": "wasm",
      "precision": "fp32",
      "fallback": false,
      "artifactRevision": "ace431824f35db231178fc602e33296784762a2e",
      "acceleration": "cpu"
    }
  },
  {
    "precision": "fp32",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "tokenParity": true,
      "winner": 0,
      "expected": 0,
      "minimumTokenCosine": 0.9999999148203532,
      "scores": [
        7.260651966428195,
        6.410420434878079,
        4.51135408977799
      ],
      "grade": "retrieval-pass"
    },
    "runtime": {
      "durationMs": 159.29999995231628,
      "inferenceMs": 159.29999995231628,
      "loadMs": 3414.5,
      "totalMs": 3573.7999999523163,
      "embeddingsPerSecond": 6.277463906461599,
      "grade": "interactive",
      "engine": "ONNX Runtime Web",
      "version": "1.24.3",
      "device": "webgpu",
      "precision": "fp32",
      "fallback": false,
      "artifactRevision": "ace431824f35db231178fc602e33296784762a2e",
      "acceleration": "partial",
      "gpuSubmissionObserved": true,
      "accelerationDetails": "WebGPU graph execution observed. Full graph GPU placement was not measured."
    }
  }
]

LateOn Code Edge 17M Pretrain

2 configurations tested
Passed2
Failed0
TestedNov 14, 2026
RuntimeONNX Runtime Web 1.24.3

Quality criteria

Three authored code-ranking fixtures, exact token IDs and agreement with original native FP32 token vectors on browser.

PrecisionDeviceResult
FP32GPU · WebGPUAcceleratedPassedVery fast
FP32CPU · WASMPassedVery fast
Test setup & sample details
Quality gate
Three authored code-ranking fixtures, exact token IDs and agreement with original native FP32 token vectors on browser.
Tested
2026-11-14T12:52:00.067Z
Environment
Apple M3 Max on macOS 27.2. Managed native or isolated Chrome module workers.
Runtime
ONNX Runtime Web 1.24.3
Artifact revision
4ca3a44b3093e72d48461aa6a67cfd5c0025c007
Fixture
{
  "file": "model-samples/lateon-code-deduplicate-query.json",
  "size": 193,
  "sha1": "aa4a1de14636660ffe22aa8f370645d65e11cace",
  "sha256": "92f8a217d38938179b0d9c1f40bab01fbdd3aba3d593558f706a7d4714205e8a",
  "url": "https://assets.localai.tools/samples/v1/files/92f8a217d38938179b0d9c1f40bab01fbdd3aba3d593558f706a7d4714205e8a/lateon-code-deduplicate-query.json",
  "durationSeconds": 0
}
Rerun when
artifact revision; runtime version; retrieval fixture; tokenizer or projection semantics
Per-case quality & failure details
[
  {
    "precision": "fp32",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "passed": true,
      "tokenParity": true,
      "winner": 0,
      "expected": 0,
      "minimumTokenCosine": 0.9999998933666262,
      "scores": [
        11.0080739233114,
        10.778504204882116,
        10.575959956649287
      ],
      "grade": "retrieval-pass"
    },
    "runtime": {
      "durationMs": 32.09999996423721,
      "inferenceMs": 32.09999996423721,
      "loadMs": 1739.300000011921,
      "totalMs": 1771.3999999761581,
      "embeddingsPerSecond": 31.152648009785217,
      "grade": "interactive",
      "engine": "ONNX Runtime Web",
      "version": "1.24.3",
      "device": "wasm",
      "precision": "fp32",
      "fallback": false,
      "artifactRevision": "4ca3a44b3093e72d48461aa6a67cfd5c0025c007",
      "acceleration": "cpu"
    }
  },
  {
    "precision": "fp32",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "tokenParity": true,
      "winner": 0,
      "expected": 0,
      "minimumTokenCosine": 0.9999998933667155,
      "scores": [
        11.008074695793713,
        10.778504296880197,
        10.57596055904585
      ],
      "grade": "retrieval-pass"
    },
    "runtime": {
      "durationMs": 96.70000004768372,
      "inferenceMs": 96.70000004768372,
      "loadMs": 1480.4000000357628,
      "totalMs": 1577.1000000834465,
      "embeddingsPerSecond": 10.34126162881996,
      "grade": "interactive",
      "engine": "ONNX Runtime Web",
      "version": "1.24.3",
      "device": "webgpu",
      "precision": "fp32",
      "fallback": false,
      "artifactRevision": "4ca3a44b3093e72d48461aa6a67cfd5c0025c007",
      "acceleration": "partial",
      "gpuSubmissionObserved": true,
      "accelerationDetails": "WebGPU graph execution observed. Full graph GPU placement was not measured."
    }
  }
]

LateOn Code Base Pretrain

2 configurations tested
Passed2
Failed0
TestedNov 14, 2026
RuntimeONNX Runtime Web 1.24.3

Quality criteria

Three authored code-ranking fixtures, exact token IDs and agreement with original native FP32 token vectors on browser.

PrecisionDeviceResult
INT8GPU · WebGPUAcceleratedPassedFast
INT8CPU · WASMPassedFast
Test setup & sample details
Quality gate
Three authored code-ranking fixtures, exact token IDs and agreement with original native FP32 token vectors on browser.
Tested
2026-11-14T12:52:20.709Z
Environment
Apple M3 Max on macOS 27.2. Managed native or isolated Chrome module workers.
Runtime
ONNX Runtime Web 1.24.3
Artifact revision
71251a6ee61eee488de7e3ae29f5fb4c3c94699b
Fixture
{
  "file": "model-samples/lateon-code-deduplicate-query.json",
  "size": 193,
  "sha1": "aa4a1de14636660ffe22aa8f370645d65e11cace",
  "sha256": "92f8a217d38938179b0d9c1f40bab01fbdd3aba3d593558f706a7d4714205e8a",
  "url": "https://assets.localai.tools/samples/v1/files/92f8a217d38938179b0d9c1f40bab01fbdd3aba3d593558f706a7d4714205e8a/lateon-code-deduplicate-query.json",
  "durationSeconds": 0
}
Rerun when
artifact revision; runtime version; retrieval fixture; tokenizer or projection semantics
Per-case quality & failure details
[
  {
    "precision": "int8",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "passed": true,
      "tokenParity": true,
      "winner": 0,
      "expected": 0,
      "minimumTokenCosine": 0.979956493992801,
      "scores": [
        10.305674701749911,
        10.235298705562382,
        9.604610369578749
      ],
      "grade": "retrieval-pass"
    },
    "runtime": {
      "durationMs": 118.30000001192093,
      "inferenceMs": 118.30000001192093,
      "loadMs": 2462.899999976158,
      "totalMs": 2581.199999988079,
      "embeddingsPerSecond": 8.453085375310494,
      "grade": "interactive",
      "engine": "ONNX Runtime Web",
      "version": "1.24.3",
      "device": "wasm",
      "precision": "int8",
      "fallback": false,
      "artifactRevision": "71251a6ee61eee488de7e3ae29f5fb4c3c94699b",
      "acceleration": "cpu"
    }
  },
  {
    "precision": "int8",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "tokenParity": true,
      "winner": 0,
      "expected": 0,
      "minimumTokenCosine": 0.9793177103759542,
      "scores": [
        10.327399781832337,
        10.231108278612057,
        9.607755891951623
      ],
      "grade": "retrieval-pass"
    },
    "runtime": {
      "durationMs": 260.89999997615814,
      "inferenceMs": 260.89999997615814,
      "loadMs": 2790.199999988079,
      "totalMs": 3051.099999964237,
      "embeddingsPerSecond": 3.8328861636312115,
      "grade": "slow",
      "engine": "ONNX Runtime Web",
      "version": "1.24.3",
      "device": "webgpu",
      "precision": "int8",
      "fallback": false,
      "artifactRevision": "71251a6ee61eee488de7e3ae29f5fb4c3c94699b",
      "acceleration": "partial",
      "gpuSubmissionObserved": true,
      "accelerationDetails": "WebGPU graph execution observed. Full graph GPU placement was not measured."
    }
  }
]

Something not quite right?

Tell us what happened with LateOn Code. 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.