--- license: apache-2.0 language: - en - es - fr - de - pt tags: - vosk - kaldi - speech-recognition pipeline_tag: automatic-speech-recognition --- # Vosk small models for Dictaria Copies of Alpha Cephei's [Vosk small models](https://alphacephei.com/vosk/models), used by [Dictaria](https://dictaria.andesprit.com) for live captions in the web browser. Each model is the original zip file repackaged as `.tar.gz`, the format [vosk-browser](https://github.com/ccoreilly/vosk-browser) reads. The model files inside are unchanged. Alpha Cephei's site does not let browsers download from it, so the app downloads these copies instead and checks each file's size and SHA-256. | File | Original model | Size (bytes) | SHA-256 | | --- | --- | --- | --- | | `vosk-model-small-en-us-0.15.tar.gz` | vosk-model-small-en-us-0.15 | 41116628 | `35c339de572da3fb8a52d6b8267c282df683f370d36a9f7a661cd447e15594dc` | | `vosk-model-small-es-0.42.tar.gz` | vosk-model-small-es-0.42 | 39748934 | `357867a8d602f1ce7f605c8ccafadf42ab441e5b34c6b55ba3c65cdad2c28fa2` | | `vosk-model-small-fr-0.22.tar.gz` | vosk-model-small-fr-0.22 | 42153811 | `7ba4da90a263a7fbafd053321b21e8c1264bd8fbef54a21171ef36965c51edaf` | | `vosk-model-small-de-0.15.tar.gz` | vosk-model-small-de-0.15 | 46376344 | `b62f0a6a93b1108cb05fdb6c3041a6a6f34c0b5c60f8f3d361c1c023ddbeff85` | | `vosk-model-small-pt-0.3.tar.gz` | vosk-model-small-pt-0.3 | 32358832 | `d7c0e5abc0a0f7457d511fd4e12bc8e9be1fc1e21893beb77189c28e5adcca4e` | ## License The models are by [Alpha Cephei](https://alphacephei.com) and are released under the Apache License 2.0, as listed on the Vosk models page.