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.

Edge0 8B A1B Preview (WebGPU)

Generate local text replies using a community Q4_K_M conversion with merged Recover-LoRA.

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

Generate local text replies using a community Q4_K_M conversion with merged Recover-LoRA.

Created by
Edge0
License
Apache-2.0
Model released
Sep 8, 2026
Source repository created
Sep 8, 2026
Model weights · Q4
4.8 GB

Best for

  • Short English and Mandarin text chat and simple follow-up questions.

Keep in mind

  • Text-only chat. This conversion uses the actual router and omits the trained prerouter.
  • Only Chrome WebGPU has measured execution. CPU WASM failed allocation.
  • Context and output share a 512-token limit. Outputs may contain mistakes.

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
https://huggingface.co/Edge0/Edge0-8B-A1B-preview
Release verification
Verified
Release evidence
Edge0's dated release announcement explicitly names both preview checkpoints.
Source updated
Oct 1, 2026
Browser model files
Proairos/Edge0-8B-A1B-preview-AideMemo-GGUF
Files updated
Sep 24, 2026
Artifact revision
17126737d742913278569000165908716101b06c

Input

CHAT & TEXT

Sample result

Edge0 8B A1B Preview (WebGPU)

Settings

Maximum number of new tokens in the reply. Higher limits allow longer answers and reasoning but take more time and memory.
Estimated speedNot benchmarked

Output

Start a conversation

Send a message or try a sample. Replies appear here.

Benchmark evidence

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

Edge0 8B A1B Preview (WebGPU)

6 configurations tested
Passed6
Failed0
TestedNov 15, 2026
Runtimewllama 3.13.0

Quality criteria

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

PrecisionDeviceResult
Q4GPU · WebGPUAcceleratedPassedVery fast
Q4GPU · WebGPUAcceleratedPassedFast
Q4GPU · WebGPUAcceleratedPassedFast
Q4GPU · WebGPUAcceleratedPassedFast
Q4GPU · WebGPUAcceleratedPassedFast
Q4GPU · WebGPUAcceleratedPassedFast
Test setup & sample details
Quality gate
No explicit quality gate recorded. Review the per-case quality evidence; full certification is pending.
Tested
2026-11-15T02:10:42.011Z
Environment
Chrome 157.0.8103.79 on Apple M3 Max macOS 27.2, actual WebGPU 25/25 layers, Wllama 3.13.0
Runtime
wllama 3.13.0
Artifact revision
17126737d742913278569000165908716101b06c
Fixture
{
  "file": "model-samples/desktop/edge0-preview/arithmetic.json",
  "size": 122,
  "sha1": "1bcd8b4fa4230e621a8b6e4801ffff3ebf366cb0",
  "sha256": "b52939c138db149890b2ff329640495a276369109975ee2556513d69c21ddf2d",
  "url": "https://assets.localai.tools/samples/v1/files/b52939c138db149890b2ff329640495a276369109975ee2556513d69c21ddf2d/arithmetic.json"
}
Rerun when
Artifact/runtime revisions; Generation/preprocessing changes; Fixture/quality scope; OS/backend behavior
Per-case quality & failure details
[
  {
    "precision": "q4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "matches": true,
      "streamingParity": true
    },
    "runtime": {
      "loadMs": 4293.899999976158,
      "inferenceMs": 1317.6000000238419,
      "totalMs": 5611.5,
      "tokensPerSecond": 21.85792349726776,
      "firstTokenMs": 1174.8999999761581,
      "grade": "short-answer",
      "provider": "Wllama WebGPU",
      "offloadedLayers": 25,
      "engine": "wllama",
      "version": "3.13.0",
      "device": "webgpu",
      "precision": "q4",
      "artifactRevision": "17126737d742913278569000165908716101b06c"
    }
  },
  {
    "precision": "q4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "matches": true,
      "streamingParity": true
    },
    "runtime": {
      "loadMs": 4293.899999976158,
      "inferenceMs": 703.1999999880791,
      "totalMs": 4997.099999964237,
      "tokensPerSecond": 54.84460695893599,
      "firstTokenMs": 619.8999999761581,
      "grade": "short-answer",
      "provider": "Wllama WebGPU",
      "offloadedLayers": 25,
      "engine": "wllama",
      "version": "3.13.0",
      "device": "webgpu",
      "precision": "q4",
      "artifactRevision": "17126737d742913278569000165908716101b06c"
    }
  },
  {
    "precision": "q4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "matches": true,
      "streamingParity": true
    },
    "runtime": {
      "loadMs": 4293.899999976158,
      "inferenceMs": 510.80000001192093,
      "totalMs": 4804.699999988079,
      "tokensPerSecond": 79.68127482471118,
      "firstTokenMs": 461,
      "grade": "short-answer",
      "provider": "Wllama WebGPU",
      "offloadedLayers": 25,
      "engine": "wllama",
      "version": "3.13.0",
      "device": "webgpu",
      "precision": "q4",
      "artifactRevision": "17126737d742913278569000165908716101b06c"
    }
  },
  {
    "precision": "q4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "matches": true,
      "streamingParity": true
    },
    "runtime": {
      "loadMs": 4293.899999976158,
      "inferenceMs": 395.30000001192093,
      "totalMs": 4689.199999988079,
      "tokensPerSecond": 87.33624440509226,
      "firstTokenMs": 349.89999997615814,
      "grade": "short-answer",
      "provider": "Wllama WebGPU",
      "offloadedLayers": 25,
      "engine": "wllama",
      "version": "3.13.0",
      "device": "webgpu",
      "precision": "q4",
      "artifactRevision": "17126737d742913278569000165908716101b06c"
    }
  },
  {
    "precision": "q4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "matches": true,
      "streamingParity": true
    },
    "runtime": {
      "loadMs": 4293.899999976158,
      "inferenceMs": 371.10000002384186,
      "totalMs": 4665,
      "tokensPerSecond": null,
      "firstTokenMs": 349.9000000357628,
      "grade": "short-answer",
      "provider": "Wllama WebGPU",
      "offloadedLayers": 25,
      "engine": "wllama",
      "version": "3.13.0",
      "device": "webgpu",
      "precision": "q4",
      "artifactRevision": "17126737d742913278569000165908716101b06c"
    }
  },
  {
    "precision": "q4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "matches": true,
      "streamingParity": true
    },
    "runtime": {
      "loadMs": 4293.899999976158,
      "inferenceMs": 598.2999999523163,
      "totalMs": 4892.199999928474,
      "tokensPerSecond": 50.61460592477676,
      "firstTokenMs": 300.0999999642372,
      "grade": "interactive",
      "provider": "Wllama WebGPU",
      "offloadedLayers": 25,
      "engine": "wllama",
      "version": "3.13.0",
      "device": "webgpu",
      "precision": "q4",
      "artifactRevision": "17126737d742913278569000165908716101b06c"
    }
  }
]

Something not quite right?

Tell us what happened with Edge0 Preview. 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.