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.

Back to models

Wav2Vec2 Base SUPERB — Emotion Recognition

Estimate neutral, happy, angry, or sad vocal patterns in English speech. Returns ranked category scores; these predictions are not reliable evidence of the speaker's actual emotional state.

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

Estimate neutral, happy, angry, or sad vocal patterns in English speech. Returns ranked category scores; these predictions are not reliable evidence of the speaker's actual emotional state.

Created by
SUPERB / S3PRL
License
Apache-2.0
Model released
Unknown
Source repository created
Mar 2, 2022
Model weights · Q4F16
66.5 MB

Best for

  • Exploratory analysis of broad vocal affect in English speech.
  • Benchmarking browser audio-classification inference.

Keep in mind

  • The four labels are coarse and do not represent the full range or ambiguity of human emotion.
  • Accuracy varies with speaker, accent, acting style, noise, recording quality, and utterance length.
  • The model was evaluated on acted English IEMOCAP speech and may not generalize to natural conversation or other languages.
  • Emotion predictions are probabilistic and must not be used for medical, employment, policing, surveillance, or other consequential judgments.

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
superb/wav2vec2-base-superb-er
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
Nov 4, 2021
Browser model files
onnx-community/wav2vec2-base-superb-er-ONNX
Files updated
Sep 3, 2026
Artifact revision
e2191e9e8b692c19c7d96e2a1d789ab1df5b7259

Input

AUDIO

Sample result

Wav2Vec2 Base SUPERB — Emotion Recognition
Record a clip, then run inference. This does not send live audio to the model.

Settings

Limits how many ranked predictions or passages are returned. Lower values show fewer top results. Higher values include weaker matches and more output. Automatic/null returns all results when offered; it does not mean the model is more accurate.

Output

Explore the tone of your audio

Choose a recording or try the sample to compare vocal emotion scores.

Benchmark evidence

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

Wav2Vec2 Base SUPERB — Emotion Recognition

10 configurations tested
Passed10
Failed0
TestedSep 12, 2026
RuntimeTransformers.js 4.2.0

Quality criteria

Finite normalized scores for four unique labels, with angry ranked first and exceeding the strongest negative by at least 0.1.

PrecisionDeviceResult
Q4F16GPU · WebGPUAcceleratedPassedVery fast
BNB4GPU · WebGPUAcceleratedPassedVery fast
Q4GPU · WebGPUAcceleratedPassedVery fast
FP32GPU · WebGPUAcceleratedPassedVery fast
FP16GPU · WebGPUAcceleratedPassedVery fast
FP16CPU · WASMPassedFast
FP32CPU · WASMPassedFast
Q4CPU · WASMPassedFast
BNB4CPU · WASMPassedFast
Q4F16CPU · WASMPassedFast
Test setup & sample details
Quality gate
Finite normalized scores for four unique labels, with angry ranked first and exceeding the strongest negative by at least 0.1.
Tested
2026-09-12T06:19:51.600Z
Environment
Not recorded
Runtime
Transformers.js 4.2.0
Artifact revision
e2191e9e8b692c19c7d96e2a1d789ab1df5b7259
Fixture
{
  "file": "crema-d-1001-dfa-ang.wav",
  "url": "https://f005.backblazeb2.com/file/hubters-webai-public/assets/benchmarks/crema-d-1001-dfa-ang.wav",
  "size": 72862,
  "sha1": "3d62067948a906c4df841268c31a96d3f9d14da5",
  "durationSeconds": 2.2755625,
  "sampleRate": 16000,
  "source": "CREMA-D 1001_DFA_ANG_XX.wav",
  "license": "ODbL-1.0 / DbCL-1.0",
  "licenseUrl": "https://github.com/CheyneyComputerScience/CREMA-D/blob/master/LICENSE.txt"
}
Rerun when
artifact revision change; runtime version change; fixture or quality-gate change; browser/backend behavior change
Per-case quality & failure details
[
  {
    "precision": "fp32",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "passed": true,
      "topLabel": "ang",
      "topScore": 0.984452,
      "margin": 0.970826
    },
    "runtime": {
      "totalMs": 37299.79999999888,
      "loadMs": 33464.9,
      "inferenceMs": 2355.7,
      "realTimeFactor": 0.345072,
      "grade": "good"
    }
  },
  {
    "precision": "fp32",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "topLabel": "ang",
      "topScore": 0.984448,
      "margin": 0.970818
    },
    "runtime": {
      "totalMs": 35471.59999999963,
      "loadMs": 33550.9,
      "inferenceMs": 494.4,
      "realTimeFactor": 0.072422,
      "grade": "excellent"
    }
  },
  {
    "precision": "fp16",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "passed": true,
      "topLabel": "ang",
      "topScore": 0.985121,
      "margin": 0.972141
    },
    "runtime": {
      "totalMs": 25224.300000000745,
      "loadMs": 21450.3,
      "inferenceMs": 2279.1,
      "realTimeFactor": 0.333852,
      "grade": "good"
    }
  },
  {
    "precision": "fp16",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "topLabel": "ang",
      "topScore": 0.9851,
      "margin": 0.972148
    },
    "runtime": {
      "totalMs": 25093.10000000149,
      "loadMs": 23096.3,
      "inferenceMs": 590.2,
      "realTimeFactor": 0.086455,
      "grade": "excellent"
    }
  },
  {
    "precision": "q4",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "passed": true,
      "topLabel": "ang",
      "topScore": 0.990658,
      "margin": 0.984401
    },
    "runtime": {
      "totalMs": 16544.700000001118,
      "loadMs": 12507.8,
      "inferenceMs": 2364.5,
      "realTimeFactor": 0.346361,
      "grade": "good"
    }
  },
  {
    "precision": "q4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "topLabel": "ang",
      "topScore": 0.990655,
      "margin": 0.984396
    },
    "runtime": {
      "totalMs": 14170.60000000149,
      "loadMs": 12514.5,
      "inferenceMs": 110.8,
      "realTimeFactor": 0.01623,
      "grade": "excellent"
    }
  },
  {
    "precision": "q4f16",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "passed": true,
      "topLabel": "ang",
      "topScore": 0.990855,
      "margin": 0.984749
    },
    "runtime": {
      "totalMs": 14627.799999998882,
      "loadMs": 10639.8,
      "inferenceMs": 2496.7,
      "realTimeFactor": 0.365726,
      "grade": "good"
    }
  },
  {
    "precision": "q4f16",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "topLabel": "ang",
      "topScore": 0.990941,
      "margin": 0.985037
    },
    "runtime": {
      "totalMs": 13422.099999999627,
      "loadMs": 11787.1,
      "inferenceMs": 87.7,
      "realTimeFactor": 0.012847,
      "grade": "excellent"
    }
  },
  {
    "precision": "bnb4",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "passed": true,
      "topLabel": "ang",
      "topScore": 0.990658,
      "margin": 0.984401
    },
    "runtime": {
      "totalMs": 17198,
      "loadMs": 13293.8,
      "inferenceMs": 2368.9,
      "realTimeFactor": 0.347006,
      "grade": "good"
    }
  },
  {
    "precision": "bnb4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "topLabel": "ang",
      "topScore": 0.990655,
      "margin": 0.984396
    },
    "runtime": {
      "totalMs": 14314.099999999627,
      "loadMs": 12688.4,
      "inferenceMs": 109,
      "realTimeFactor": 0.015967,
      "grade": "excellent"
    }
  }
]

Something not quite right?

Tell us what happened with Wav2Vec2 Emotion Recognition. 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.