3c:I[603186,["/_next/static/chunks/1176bnf8oism6.js","/_next/static/chunks/0aiio4ivp1n07.js","/_next/static/chunks/3cauv554e_gxu.js","/_next/static/chunks/3byf1h7-4uxfn.js","/_next/static/chunks/3r5yd4k4rms8-.js","/_next/static/chunks/1la1pzs9j8idc.js","/_next/static/chunks/12haym2zhwrql.js","/_next/static/chunks/0gvbio555n823.js","/_next/static/chunks/3ygl6_80ds9-d.js","/_next/static/chunks/2m_mi71579pjv.js","/_next/static/chunks/3pjc6r7w9qxan.js","/_next/static/chunks/07t1rcdzji0vn.js","/_next/static/chunks/0ab2-46782ntx.js","/_next/static/chunks/3ex1ijq1gw1eb.js","/_next/static/chunks/27f80-sq8a9nm.js","/_next/static/chunks/21k90yd6lrzsc.js","/_next/static/chunks/00-5kse4g6azx.js","/_next/static/chunks/3giy7jirffwz9.js","/_next/static/chunks/2vsadn5_3a9ui.js","/_next/static/chunks/0bvim9_g0hu68.js","/_next/static/chunks/0gf5i4fmb2vxz.js","/_next/static/chunks/2bcd-9k-4fyab.js","/_next/static/chunks/18gle79tm4_vm.js","/_next/static/chunks/23c-cbn9bzzbb.js","/_next/static/chunks/1_lyn-lih218f.js","/_next/static/chunks/1c0w6764iswod.js","/_next/static/chunks/3d865g9_und-a.js","/_next/static/chunks/0tt83uyswwa6b.js","/_next/static/chunks/0czpc-f8wjri8.js","/_next/static/chunks/24_gg9g_4g22i.js","/_next/static/chunks/31pwnoicz230f.js","/_next/static/chunks/2dyt8rg18m736.js","/_next/static/chunks/312elt506g3__.js","/_next/static/chunks/0n-s3bbshxp3e.js","/_next/static/chunks/2br2z5e4g144b.js","/_next/static/chunks/2pqhypz2vo8mu.js","/_next/static/chunks/2fz52lc6kg1s_.js","/_next/static/chunks/3x3-h1iyted1x.js","/_next/static/chunks/0fpx6rmsgi4um.js"],"PageMetadata"] 3d:I[831915,["/_next/static/chunks/1176bnf8oism6.js","/_next/static/chunks/0aiio4ivp1n07.js","/_next/static/chunks/3cauv554e_gxu.js","/_next/static/chunks/3byf1h7-4uxfn.js","/_next/static/chunks/3r5yd4k4rms8-.js","/_next/static/chunks/1la1pzs9j8idc.js","/_next/static/chunks/12haym2zhwrql.js","/_next/static/chunks/0gvbio555n823.js","/_next/static/chunks/3ygl6_80ds9-d.js","/_next/static/chunks/2m_mi71579pjv.js","/_next/static/chunks/3pjc6r7w9qxan.js","/_next/static/chunks/07t1rcdzji0vn.js","/_next/static/chunks/0ab2-46782ntx.js","/_next/static/chunks/3ex1ijq1gw1eb.js","/_next/static/chunks/27f80-sq8a9nm.js","/_next/static/chunks/21k90yd6lrzsc.js","/_next/static/chunks/00-5kse4g6azx.js","/_next/static/chunks/3giy7jirffwz9.js","/_next/static/chunks/2vsadn5_3a9ui.js","/_next/static/chunks/0bvim9_g0hu68.js","/_next/static/chunks/0gf5i4fmb2vxz.js","/_next/static/chunks/2bcd-9k-4fyab.js","/_next/static/chunks/18gle79tm4_vm.js","/_next/static/chunks/23c-cbn9bzzbb.js","/_next/static/chunks/1_lyn-lih218f.js","/_next/static/chunks/1c0w6764iswod.js","/_next/static/chunks/3d865g9_und-a.js","/_next/static/chunks/0tt83uyswwa6b.js","/_next/static/chunks/0czpc-f8wjri8.js","/_next/static/chunks/24_gg9g_4g22i.js","/_next/static/chunks/31pwnoicz230f.js","/_next/static/chunks/2dyt8rg18m736.js","/_next/static/chunks/312elt506g3__.js","/_next/static/chunks/0n-s3bbshxp3e.js","/_next/static/chunks/2br2z5e4g144b.js","/_next/static/chunks/2pqhypz2vo8mu.js","/_next/static/chunks/2fz52lc6kg1s_.js","/_next/static/chunks/3x3-h1iyted1x.js","/_next/static/chunks/0fpx6rmsgi4um.js"],"SearchResultsImage"] 3e:I[66106,["/_next/static/chunks/1176bnf8oism6.js","/_next/static/chunks/0aiio4ivp1n07.js","/_next/static/chunks/3cauv554e_gxu.js","/_next/static/chunks/3byf1h7-4uxfn.js","/_next/static/chunks/3r5yd4k4rms8-.js","/_next/static/chunks/1la1pzs9j8idc.js","/_next/static/chunks/12haym2zhwrql.js","/_next/static/chunks/0gvbio555n823.js","/_next/static/chunks/3ygl6_80ds9-d.js","/_next/static/chunks/2m_mi71579pjv.js","/_next/static/chunks/3pjc6r7w9qxan.js","/_next/static/chunks/07t1rcdzji0vn.js","/_next/static/chunks/0ab2-46782ntx.js","/_next/static/chunks/3ex1ijq1gw1eb.js","/_next/static/chunks/27f80-sq8a9nm.js","/_next/static/chunks/21k90yd6lrzsc.js","/_next/static/chunks/00-5kse4g6azx.js","/_next/static/chunks/3giy7jirffwz9.js","/_next/static/chunks/2vsadn5_3a9ui.js","/_next/static/chunks/0bvim9_g0hu68.js","/_next/static/chunks/0gf5i4fmb2vxz.js","/_next/static/chunks/2bcd-9k-4fyab.js","/_next/static/chunks/18gle79tm4_vm.js","/_next/static/chunks/23c-cbn9bzzbb.js","/_next/static/chunks/1_lyn-lih218f.js","/_next/static/chunks/1c0w6764iswod.js","/_next/static/chunks/3d865g9_und-a.js","/_next/static/chunks/0tt83uyswwa6b.js","/_next/static/chunks/0czpc-f8wjri8.js","/_next/static/chunks/24_gg9g_4g22i.js","/_next/static/chunks/31pwnoicz230f.js","/_next/static/chunks/2dyt8rg18m736.js","/_next/static/chunks/312elt506g3__.js","/_next/static/chunks/0n-s3bbshxp3e.js","/_next/static/chunks/2br2z5e4g144b.js","/_next/static/chunks/2pqhypz2vo8mu.js","/_next/static/chunks/2fz52lc6kg1s_.js","/_next/static/chunks/3x3-h1iyted1x.js","/_next/static/chunks/0fpx6rmsgi4um.js"],""] 3f:I[807982,["/_next/static/chunks/1176bnf8oism6.js","/_next/static/chunks/0aiio4ivp1n07.js","/_next/static/chunks/3cauv554e_gxu.js","/_next/static/chunks/3byf1h7-4uxfn.js","/_next/static/chunks/3r5yd4k4rms8-.js","/_next/static/chunks/1la1pzs9j8idc.js","/_next/static/chunks/12haym2zhwrql.js","/_next/static/chunks/0gvbio555n823.js","/_next/static/chunks/3ygl6_80ds9-d.js","/_next/static/chunks/2m_mi71579pjv.js","/_next/static/chunks/3pjc6r7w9qxan.js","/_next/static/chunks/07t1rcdzji0vn.js","/_next/static/chunks/0ab2-46782ntx.js","/_next/static/chunks/3ex1ijq1gw1eb.js","/_next/static/chunks/27f80-sq8a9nm.js","/_next/static/chunks/21k90yd6lrzsc.js","/_next/static/chunks/00-5kse4g6azx.js","/_next/static/chunks/3giy7jirffwz9.js","/_next/static/chunks/2vsadn5_3a9ui.js","/_next/static/chunks/0bvim9_g0hu68.js","/_next/static/chunks/0gf5i4fmb2vxz.js","/_next/static/chunks/2bcd-9k-4fyab.js","/_next/static/chunks/18gle79tm4_vm.js","/_next/static/chunks/23c-cbn9bzzbb.js","/_next/static/chunks/1_lyn-lih218f.js","/_next/static/chunks/1c0w6764iswod.js","/_next/static/chunks/3d865g9_und-a.js","/_next/static/chunks/0tt83uyswwa6b.js","/_next/static/chunks/0czpc-f8wjri8.js","/_next/static/chunks/24_gg9g_4g22i.js","/_next/static/chunks/31pwnoicz230f.js","/_next/static/chunks/2dyt8rg18m736.js","/_next/static/chunks/312elt506g3__.js","/_next/static/chunks/0n-s3bbshxp3e.js","/_next/static/chunks/2br2z5e4g144b.js","/_next/static/chunks/2pqhypz2vo8mu.js","/_next/static/chunks/2fz52lc6kg1s_.js","/_next/static/chunks/3x3-h1iyted1x.js","/_next/static/chunks/0fpx6rmsgi4um.js"],"default"] 40:I[299632,["/_next/static/chunks/1176bnf8oism6.js","/_next/static/chunks/0aiio4ivp1n07.js","/_next/static/chunks/3cauv554e_gxu.js","/_next/static/chunks/3byf1h7-4uxfn.js","/_next/static/chunks/3r5yd4k4rms8-.js","/_next/static/chunks/1la1pzs9j8idc.js","/_next/static/chunks/12haym2zhwrql.js","/_next/static/chunks/0gvbio555n823.js","/_next/static/chunks/3ygl6_80ds9-d.js","/_next/static/chunks/2m_mi71579pjv.js","/_next/static/chunks/3pjc6r7w9qxan.js","/_next/static/chunks/07t1rcdzji0vn.js","/_next/static/chunks/0ab2-46782ntx.js","/_next/static/chunks/3ex1ijq1gw1eb.js","/_next/static/chunks/27f80-sq8a9nm.js","/_next/static/chunks/21k90yd6lrzsc.js","/_next/static/chunks/00-5kse4g6azx.js","/_next/static/chunks/3giy7jirffwz9.js","/_next/static/chunks/2vsadn5_3a9ui.js","/_next/static/chunks/0bvim9_g0hu68.js","/_next/static/chunks/0gf5i4fmb2vxz.js","/_next/static/chunks/2bcd-9k-4fyab.js","/_next/static/chunks/18gle79tm4_vm.js","/_next/static/chunks/23c-cbn9bzzbb.js","/_next/static/chunks/1_lyn-lih218f.js","/_next/static/chunks/1c0w6764iswod.js","/_next/static/chunks/3d865g9_und-a.js","/_next/static/chunks/0tt83uyswwa6b.js","/_next/static/chunks/0czpc-f8wjri8.js","/_next/static/chunks/24_gg9g_4g22i.js","/_next/static/chunks/31pwnoicz230f.js","/_next/static/chunks/2dyt8rg18m736.js","/_next/static/chunks/312elt506g3__.js","/_next/static/chunks/0n-s3bbshxp3e.js","/_next/static/chunks/2br2z5e4g144b.js","/_next/static/chunks/2pqhypz2vo8mu.js","/_next/static/chunks/2fz52lc6kg1s_.js","/_next/static/chunks/3x3-h1iyted1x.js","/_next/static/chunks/0fpx6rmsgi4um.js"],"default"] 41:I[400402,["/_next/static/chunks/1176bnf8oism6.js","/_next/static/chunks/0aiio4ivp1n07.js","/_next/static/chunks/3cauv554e_gxu.js","/_next/static/chunks/3byf1h7-4uxfn.js","/_next/static/chunks/3r5yd4k4rms8-.js","/_next/static/chunks/1la1pzs9j8idc.js","/_next/static/chunks/12haym2zhwrql.js","/_next/static/chunks/0gvbio555n823.js","/_next/static/chunks/3ygl6_80ds9-d.js","/_next/static/chunks/2m_mi71579pjv.js","/_next/static/chunks/3pjc6r7w9qxan.js","/_next/static/chunks/07t1rcdzji0vn.js","/_next/static/chunks/0ab2-46782ntx.js","/_next/static/chunks/3ex1ijq1gw1eb.js","/_next/static/chunks/27f80-sq8a9nm.js","/_next/static/chunks/21k90yd6lrzsc.js","/_next/static/chunks/00-5kse4g6azx.js","/_next/static/chunks/3giy7jirffwz9.js","/_next/static/chunks/2vsadn5_3a9ui.js","/_next/static/chunks/0bvim9_g0hu68.js","/_next/static/chunks/0gf5i4fmb2vxz.js","/_next/static/chunks/2bcd-9k-4fyab.js","/_next/static/chunks/18gle79tm4_vm.js","/_next/static/chunks/23c-cbn9bzzbb.js","/_next/static/chunks/1_lyn-lih218f.js","/_next/static/chunks/1c0w6764iswod.js","/_next/static/chunks/3d865g9_und-a.js","/_next/static/chunks/0tt83uyswwa6b.js","/_next/static/chunks/0czpc-f8wjri8.js","/_next/static/chunks/24_gg9g_4g22i.js","/_next/static/chunks/31pwnoicz230f.js","/_next/static/chunks/2dyt8rg18m736.js","/_next/static/chunks/312elt506g3__.js","/_next/static/chunks/0n-s3bbshxp3e.js","/_next/static/chunks/2br2z5e4g144b.js","/_next/static/chunks/2pqhypz2vo8mu.js","/_next/static/chunks/2fz52lc6kg1s_.js","/_next/static/chunks/3x3-h1iyted1x.js","/_next/static/chunks/0fpx6rmsgi4um.js"],"EndpointCTA"] 42:I[331699,["/_next/static/chunks/1176bnf8oism6.js","/_next/static/chunks/0aiio4ivp1n07.js","/_next/static/chunks/3cauv554e_gxu.js","/_next/static/chunks/3byf1h7-4uxfn.js","/_next/static/chunks/3r5yd4k4rms8-.js","/_next/static/chunks/1la1pzs9j8idc.js","/_next/static/chunks/12haym2zhwrql.js","/_next/static/chunks/0gvbio555n823.js","/_next/static/chunks/3ygl6_80ds9-d.js","/_next/static/chunks/2m_mi71579pjv.js","/_next/static/chunks/3pjc6r7w9qxan.js","/_next/static/chunks/07t1rcdzji0vn.js","/_next/static/chunks/0ab2-46782ntx.js","/_next/static/chunks/3ex1ijq1gw1eb.js","/_next/static/chunks/27f80-sq8a9nm.js","/_next/static/chunks/21k90yd6lrzsc.js","/_next/static/chunks/00-5kse4g6azx.js","/_next/static/chunks/3giy7jirffwz9.js","/_next/static/chunks/2vsadn5_3a9ui.js","/_next/static/chunks/0bvim9_g0hu68.js","/_next/static/chunks/0gf5i4fmb2vxz.js","/_next/static/chunks/2bcd-9k-4fyab.js","/_next/static/chunks/18gle79tm4_vm.js","/_next/static/chunks/23c-cbn9bzzbb.js","/_next/static/chunks/1_lyn-lih218f.js","/_next/static/chunks/1c0w6764iswod.js","/_next/static/chunks/3d865g9_und-a.js","/_next/static/chunks/0tt83uyswwa6b.js","/_next/static/chunks/0czpc-f8wjri8.js","/_next/static/chunks/24_gg9g_4g22i.js","/_next/static/chunks/31pwnoicz230f.js","/_next/static/chunks/2dyt8rg18m736.js","/_next/static/chunks/312elt506g3__.js","/_next/static/chunks/0n-s3bbshxp3e.js","/_next/static/chunks/2br2z5e4g144b.js","/_next/static/chunks/2pqhypz2vo8mu.js","/_next/static/chunks/2fz52lc6kg1s_.js","/_next/static/chunks/3x3-h1iyted1x.js","/_next/static/chunks/0fpx6rmsgi4um.js"],"ModelContent"] 48:I[227209,["/_next/static/chunks/1176bnf8oism6.js","/_next/static/chunks/0aiio4ivp1n07.js","/_next/static/chunks/3cauv554e_gxu.js","/_next/static/chunks/3byf1h7-4uxfn.js","/_next/static/chunks/3r5yd4k4rms8-.js","/_next/static/chunks/1la1pzs9j8idc.js","/_next/static/chunks/12haym2zhwrql.js","/_next/static/chunks/0gvbio555n823.js","/_next/static/chunks/3ygl6_80ds9-d.js","/_next/static/chunks/2m_mi71579pjv.js","/_next/static/chunks/3pjc6r7w9qxan.js","/_next/static/chunks/07t1rcdzji0vn.js","/_next/static/chunks/0ab2-46782ntx.js","/_next/static/chunks/3ex1ijq1gw1eb.js","/_next/static/chunks/27f80-sq8a9nm.js","/_next/static/chunks/21k90yd6lrzsc.js","/_next/static/chunks/00-5kse4g6azx.js","/_next/static/chunks/3giy7jirffwz9.js","/_next/static/chunks/2vsadn5_3a9ui.js","/_next/static/chunks/0bvim9_g0hu68.js","/_next/static/chunks/0gf5i4fmb2vxz.js","/_next/static/chunks/2bcd-9k-4fyab.js","/_next/static/chunks/18gle79tm4_vm.js","/_next/static/chunks/23c-cbn9bzzbb.js","/_next/static/chunks/1_lyn-lih218f.js","/_next/static/chunks/1c0w6764iswod.js","/_next/static/chunks/3d865g9_und-a.js","/_next/static/chunks/0tt83uyswwa6b.js","/_next/static/chunks/0czpc-f8wjri8.js","/_next/static/chunks/24_gg9g_4g22i.js","/_next/static/chunks/31pwnoicz230f.js","/_next/static/chunks/2dyt8rg18m736.js","/_next/static/chunks/312elt506g3__.js","/_next/static/chunks/0n-s3bbshxp3e.js","/_next/static/chunks/2br2z5e4g144b.js","/_next/static/chunks/2pqhypz2vo8mu.js","/_next/static/chunks/2fz52lc6kg1s_.js","/_next/static/chunks/3x3-h1iyted1x.js","/_next/static/chunks/0fpx6rmsgi4um.js"],"default"] 49:I[946424,["/_next/static/chunks/1176bnf8oism6.js","/_next/static/chunks/0aiio4ivp1n07.js","/_next/static/chunks/3cauv554e_gxu.js","/_next/static/chunks/3byf1h7-4uxfn.js","/_next/static/chunks/3r5yd4k4rms8-.js","/_next/static/chunks/1la1pzs9j8idc.js","/_next/static/chunks/12haym2zhwrql.js","/_next/static/chunks/0gvbio555n823.js","/_next/static/chunks/3ygl6_80ds9-d.js","/_next/static/chunks/2m_mi71579pjv.js","/_next/static/chunks/3pjc6r7w9qxan.js","/_next/static/chunks/07t1rcdzji0vn.js","/_next/static/chunks/0ab2-46782ntx.js","/_next/static/chunks/3ex1ijq1gw1eb.js","/_next/static/chunks/27f80-sq8a9nm.js","/_next/static/chunks/21k90yd6lrzsc.js","/_next/static/chunks/00-5kse4g6azx.js","/_next/static/chunks/3giy7jirffwz9.js","/_next/static/chunks/2vsadn5_3a9ui.js","/_next/static/chunks/0bvim9_g0hu68.js","/_next/static/chunks/0gf5i4fmb2vxz.js","/_next/static/chunks/2bcd-9k-4fyab.js","/_next/static/chunks/18gle79tm4_vm.js","/_next/static/chunks/23c-cbn9bzzbb.js","/_next/static/chunks/1_lyn-lih218f.js","/_next/static/chunks/1c0w6764iswod.js","/_next/static/chunks/3d865g9_und-a.js","/_next/static/chunks/0tt83uyswwa6b.js","/_next/static/chunks/0czpc-f8wjri8.js","/_next/static/chunks/24_gg9g_4g22i.js","/_next/static/chunks/31pwnoicz230f.js","/_next/static/chunks/2dyt8rg18m736.js","/_next/static/chunks/312elt506g3__.js","/_next/static/chunks/0n-s3bbshxp3e.js","/_next/static/chunks/2br2z5e4g144b.js","/_next/static/chunks/2pqhypz2vo8mu.js","/_next/static/chunks/2fz52lc6kg1s_.js","/_next/static/chunks/3x3-h1iyted1x.js","/_next/static/chunks/0fpx6rmsgi4um.js"],"default"] 39:["$","$L3c",null,{"hasPartnerEndpoint":false}] 43:T3799,**Parakeet-tdt-0.6b-v3** is a 600-million-parameter multilingual automatic speech recognition (ASR) model designed for high-throughput speech-to-text transcription. It extends the parakeet-tdt-0.6b-v2 model by expanding language support from English to 25 European languages. The model automatically detects the language of the audio and transcribes it without requiring additional prompting. It is part of a series of models that leverage the Granary \[1, 2\] multilingual corpus as their primary training dataset. Key Features: * Automatic language detection across 25 European languages * Automatic punctuation and capitalization * Accurate word-level and segment-level timestamps * Long audio transcription, supporting audio up to 24 minutes long with full attention (on A100 80GB) or up to 3 hours with local attention. Try the experience here: [https://huggingface.co/spaces/nvidia/parakeet-tdt-0.6b-v3](https://huggingface.co/spaces/nvidia/parakeet-tdt-0.6b-v3) Supported Languages: Bulgarian (bg), Croatian (hr), Czech (cs), Danish (da), Dutch (nl), English (en), Estonian (et), Finnish (fi), French (fr), German (de), Greek (el), Hungarian (hu), Italian (it), Latvian (lv), Lithuanian (lt), Maltese (mt), Polish (pl), Portuguese (pt), Romanian (ro), Slovak (sk), Slovenian (sl), Spanish (es), Swedish (sv), Russian (ru), Ukrainian (uk) **GOVERNING TERMS:** Use of this **Parakeet-tdt-0.6b-v3** is governed by the NVIDIA open model licence agreement (found at [NVIDIA Open Model License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)). ## Automatic Speech Recognition (ASR) Performance ![ASR WER Comparison](https://assets.ngc.nvidia.com/products/api-catalog/parakeet-tdt-0_6b/asr.jpg) *Figure 1: ASR WER comparison across different models. This does not include Punctuation and Capitalisation errors.* --- ### Evaluation Notes **Note 1:** The above evaluations are conducted for 24 supported languages, excluding Latvian since `seamless-m4t-v2-large` and `seamless-m4t-medium` do not support it. **Note 2:** Performance differences may be partly attributed to Portuguese variant differences - our training data uses European Portuguese while most benchmarks use Brazilian Portuguese. ### Deployment Geography: Global ### Use Case: This model serves developers, researchers, academics, and industries building applications that require speech-to-text capabilities, including but not limited to: conversational AI, voice assistants, transcription services, subtitle generation, and voice analytics platforms. ### Release Date: Huggingface [08/14/2025](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3) ### Model Architecture: **Architecture Type**: FastConformer-TDT **Network Architecture**: * This model was developed based on [FastConformer encoder](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/main/asr/models.html#fast-conformer) architecture[3] and TDT decoder[4] * This model has 600 million model parameters. ### Input: **Input Type(s):** 16kHz Audio **Input Format(s):** `.wav` and `.flac` audio formats **Input Parameters:** 1D (audio signal) **Other Properties Related to Input:** Monochannel audio ### Output: **Output Type(s):** Text **Output Format:** String **Output Parameters:** 1D (text) **Other Properties Related to Output:** Punctuations and Capitalizations included. Our AI models are designed and/or optimized to run on NVIDIA GPU-accelerated systems. By leveraging NVIDIA's hardware (e.g. GPU cores) and software frameworks (e.g., CUDA libraries), the model achieves faster training and inference times compared to CPU-only solutions. For more information, refer to the [NeMo documentation](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/main/asr/models.html#fast-conformer). ## How to Use this Model: To train, fine-tune or play with the model you will need to install [NVIDIA NeMo](https://github.com/NVIDIA/NeMo). We recommend you install it after you've installed latest PyTorch version. ```bash pip install -U nemo_toolkit['asr'] ``` The model is available for use in the NeMo toolkit [5], and can be used as a pre-trained checkpoint for inference or for fine-tuning on another dataset. #### Automatically instantiate the model ```python import nemo.collections.asr as nemo_asr asr_model = nemo_asr.models.ASRModel.from_pretrained(model_name="nvidia/parakeet-tdt-0.6b-v3") ``` #### Transcribing using Python First, let's get a sample ```bash wget https://dldata-public.s3.us-east-2.amazonaws.com/2086-149220-0033.wav ``` Then simply do: ```python output = asr_model.transcribe(['2086-149220-0033.wav']) print(output[0].text) ``` #### Transcribing with timestamps To transcribe with timestamps: ```python output = asr_model.transcribe(['2086-149220-0033.wav'], timestamps=True) # by default, timestamps are enabled for char, word and segment level word_timestamps = output[0].timestamp['word'] # word level timestamps for first sample segment_timestamps = output[0].timestamp['segment'] # segment level timestamps char_timestamps = output[0].timestamp['char'] # char level timestamps for stamp in segment_timestamps: print(f"{stamp['start']}s - {stamp['end']}s : {stamp['segment']}") ``` #### Transcribing long-form audio ```python #updating self-attention model of fast-conformer encoder #setting attention left and right context sizes to 256 asr_model.change_attention_model(self_attention_model="rel_pos_local_attn", att_context_size=[256, 256]) output = asr_model.transcribe(['2086-149220-0033.wav']) print(output[0].text) ``` ## Software Integration: **Runtime Engine(s):** * NeMo 2.2 **Supported Hardware Microarchitecture Compatibility:** * NVIDIA Ampere * NVIDIA Blackwell * NVIDIA Hopper * NVIDIA Volta **[Preferred/Supported] Operating System(s):** - Linux **Hardware Specific Requirements:** Atleast 2GB RAM for model to load. The bigger the RAM, the larger audio input it supports. #### Model Version Current version: `parakeet-tdt-0.6b-v3`. Previous versions can be [accessed](https://huggingface.co/collections/nvidia/parakeet-659711f49d1469e51546e021) here. ## Training and Evaluation Datasets: ### Training This model was trained using the NeMo toolkit [5], following the strategies below: - Initialized from a CTC multilingual checkpoint pretrained on the Granary dataset \[1] \[2]. - Trained for 150,000 steps on 128 A100 GPUs. - Dataset corpora and languages were balanced using a temperature sampling value of 0.5. - Stage 2 fine-tuning was performed for 5,000 steps on 4 A100 GPUs using approximately 7,500 hours of high-quality, human-transcribed data of NeMo ASR Set 3.0. Training was conducted using this [example script](https://github.com/NVIDIA/NeMo/blob/main/examples/asr/asr_transducer/speech_to_text_rnnt_bpe.py) and [TDT configuration](https://github.com/NVIDIA/NeMo/blob/main/examples/asr/conf/fastconformer/hybrid_transducer_ctc/fastconformer_hybrid_tdt_ctc_bpe.yaml). During the training, a unified SentencePiece Tokenizer \[6] with a vocabulary of **8,192 tokens** was used. The unified tokenizer was constructed from the training set transcripts using this [script](https://github.com/NVIDIA/NeMo/blob/main/scripts/tokenizers/process_asr_text_tokenizer.py) and was optimized across all 25 supported languages. ### Training Dataset The model was trained on the combination of [Granary dataset's ASR subset](https://huggingface.co/datasets/nvidia/Granary) and in-house dataset NeMo ASR Set 3.0: - 10,000 hours from human-transcribed NeMo ASR Set 3.0, including: - LibriSpeech (960 hours) - Fisher Corpus - National Speech Corpus Part 1 - VCTK - Europarl-ASR - Multilingual LibriSpeech - Mozilla Common Voice (v7.0) - AMI - 660,000 hours of pseudo-labeled data from Granary \[1] \[2], including: - [YTC](https://huggingface.co/datasets/PleIAs/YouTube-Commons) \[7] - [MOSEL](https://huggingface.co/datasets/FBK-MT/mosel) \[8] - [YODAS](https://huggingface.co/datasets/espnet/yodas-granary) \[9] All transcriptions preserve punctuation and capitalization. The Granary dataset will be made publicly available after presentation at Interspeech 2025. **Data Collection Method by dataset** * Hybrid: Automated, Human **Labeling Method by dataset** * Hybrid: Synthetic, Human **Properties:** * Noise robust data from various sources * Single channel, 16kHz sampled data #### Evaluation Datasets For multilingual ASR performance evaluation: - Fleurs [10] - MLS [11] - CoVoST [12] For English ASR performance evaluation: - Hugging Face Open ASR Leaderboard [13] datasets **Data Collection Method by dataset** * Human **Labeling Method by dataset** * Human **Properties:** * All are commonly used for benchmarking English ASR systems. * Audio data is typically processed into a 16kHz mono channel format for ASR evaluation, consistent with benchmarks like the [Open ASR Leaderboard](https://huggingface.co/spaces/hf-audio/open_asr_leaderboard). ## Performance #### Multilingual ASR The tables below summarizes the WER (%) using a Transducer decoder with greedy decoding (without an external language model): | Language | Fleurs | MLS | CoVoST | |----------|--------|-----|--------| | **Average WER ↓** | *11.97%* | *7.83%* | *11.98%* | | **bg** | 12.64% | - | - | | **cs** | 11.01% | - | - | | **da** | 18.41% | - | - | | **de** | 5.04% | - | 4.84% | | **el** | 20.70% | - | - | | **en** | 4.85% | - | 6.80% | | **es** | 3.45% | 4.39% | 3.41% | | **et** | 17.73% | - | 22.04% | | **fi** | 13.21% | - | - | | **fr** | 5.15% | 4.97% | 6.05% | | **hr** | 12.46% | - | - | | **hu** | 15.72% | - | - | | **it** | 3.00% | 10.08% | 3.69% | | **lt** | 20.35% | - | - | | **lv** | 22.84% | - | 38.36% | | **mt** | 20.46% | - | - | | **nl** | 7.48% | 12.78% | 6.50% | | **pl** | 7.31% | 7.28% | - | | **pt** | 4.76% | 7.50% | 3.96% | | **ro** | 12.44% | - | - | | **ru** | 5.51% | - | 3.00% | | **sk** | 8.82% | - | - | | **sl** | 24.03% | - | 31.80% | | **sv** | 15.08% | - | 20.16% | | **uk** | 6.79% | - | 5.10% | **Note:** WERs are calculated after removing Punctuation and Capitalization from reference and predicted text. #### Huggingface Open-ASR-Leaderboard | **Model** | **Avg WER** | **AMI** | **Earnings-22** | **GigaSpeech** | **LS test-clean** | **LS test-other** | **SPGI Speech** | **TEDLIUM-v3** | **VoxPopuli** | |:-------------|:-------------:|:---------:|:------------------:|:----------------:|:-----------------:|:-----------------:|:------------------:|:----------------:|:---------------:| | `parakeet-tdt-0.6b-v3` | 6.34% | 11.31% | 11.42% | 9.59% | 1.93% | 3.59% | 3.97% | 2.75% | 6.14% | Additional evaluation details are available on the [Hugging Face ASR Leaderboard](https://huggingface.co/spaces/hf-audio/open_asr_leaderboard).[13] ### Noise Robustness Performance across different Signal-to-Noise Ratios (SNR) using MUSAN music and noise samples [14]: | **SNR Level** | **Avg WER** | **AMI** | **Earnings** | **GigaSpeech** | **LS test-clean** | **LS test-other** | **SPGI** | **Tedlium** | **VoxPopuli** | **Relative Change** | |:---------------|:-------------:|:----------:|:------------:|:----------------:|:-----------------:|:-----------------:|:-----------:|:-------------:|:---------------:|:-----------------:| | Clean | 6.34% | 11.31% | 11.42% | 9.59% | 1.93% | 3.59% | 3.97% | 2.75% | 6.14% | - | | SNR 10 | 7.12% | 13.99% | 11.79% | 9.96% | 2.15% | 4.55% | 4.45% | 3.05% | 6.99% | -12.28% | | SNR 5 | 8.23% | 17.59% | 13.01% | 10.69% | 2.62% | 6.05% | 5.23% | 3.33% | 7.31% | -29.81% | | SNR 0 | 11.66% | 24.44% | 17.34% | 13.60% | 4.82% | 10.38% | 8.41% | 5.39% | 8.91% | -83.97% | | SNR -5 | 19.88% | 34.91% | 26.92% | 21.41% | 12.21% | 19.98% | 16.96% | 11.36% | 15.30% | -213.64% | ## References [1] [Granary: Speech Recognition and Translation Dataset in 25 European Languages](https://arxiv.org/abs/2505.13404) [2] [NVIDIA Granary Dataset Card](https://huggingface.co/datasets/nvidia/Granary) [3] [Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition](https://arxiv.org/abs/2305.05084) [4] [Efficient Sequence Transduction by Jointly Predicting Tokens and Durations](https://arxiv.org/abs/2304.06795) [5] [NVIDIA NeMo Toolkit](https://github.com/NVIDIA/NeMo) [6] [Google Sentencepiece Tokenizer](https://github.com/google/sentencepiece) [7] [Youtube-Commons](https://huggingface.co/datasets/PleIAs/YouTube-Commons) [8] [MOSEL: 950,000 Hours of Speech Data for Open-Source Speech Foundation Model Training on EU Languages](https://arxiv.org/abs/2410.01036) [9] [YODAS: Youtube-Oriented Dataset for Audio and Speech](https://arxiv.org/pdf/2406.00899) [10] [FLEURS: Few-shot Learning Evaluation of Universal Representations of Speech](https://arxiv.org/abs/2205.12446) [11] [MLS: A Large-Scale Multilingual Dataset for Speech Research](https://arxiv.org/abs/2012.03411) [12] [CoVoST 2 and Massively Multilingual Speech-to-Text Translation](https://arxiv.org/abs/2007.10310) [13] [HuggingFace ASR Leaderboard](https://huggingface.co/spaces/hf-audio/open_asr_leaderboard) [14] [MUSAN: A Music, Speech, and Noise Corpus](https://arxiv.org/abs/1510.08484) ## Inference: **Engine**: * NVIDIA NeMo **Test Hardware**: * NVIDIA A10 * NVIDIA A100 * NVIDIA A30 * NVIDIA H100 * NVIDIA L4 * NVIDIA L40 * NVIDIA Turing T4 * NVIDIA Volta V100 ## Ethical Considerations: NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for this model, please see the Model Card++ Explainability, Bias, Safety & Security, and Privacy Subcards [here](https://developer.nvidia.com/blog/enhancing-ai-transparency-and-ethical-considerations-with-model-card/). Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).44:T930,Field | Response ------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------- Intended Domain | Speech to Text Transcription Model Type | FastConformer Intended Users | This model is intended for developers, researchers, academics, and industries building conversational based applications. Output | Text Describe how the model works | Speech input is encoded into embeddings and passed into conformer-based model and output a text response. Name the adversely impacted groups this has been tested to deliver comparable outcomes regardless of | Not Applicable Technical Limitations & Mitigation | Transcripts may be not 100% accurate. Accuracy varies based on language and characteristics of input audio (Domain, Use Case, Accent, Noise, Speech Type, Context of speech, etc.) Verified to have met prescribed NVIDIA quality standards | Yes Performance Metrics | Word Error Rate Potential Known Risks | If a word is not trained in the language model and not presented in vocabulary, the word is not likely to be recognized. Not recommended for word-for-word/incomplete sentences as accuracy varies based on the context of input text Licensing | GOVERNING TERMS: Use of this model is governed by the NVIDIA API Trial Terms of Service.45:T460,Field | Response ----------------------------------------------------------------------------------------------------------------------------------|----------------------------------------------- Generatable or reverse engineerable personal data? | None Personal data used to create this model? | None Is there provenance for all datasets used in training? | Yes Does data labeling (annotation, metadata) comply with privacy laws? | Yes Is data compliant with data subject requests for data correction or removal, if such a request was made? | No, not possible with externally-sourced data. Applicable Privacy Policy | https://www.nvidia.com/en-us/about-nvidia/privacy-policy/46:Ta26, Riva supports gRPC, HTTP and Realtime APIs. ## Inference using HTTP API Inference using HTTP API. Replace `` with the full path of the wav file to be transcribed. This model serves 25 European locales — pass the BCP-47 code matching your audio (e.g. `en-GB`, `de-DE`, `fr-FR`, `es-ES`). ```bash curl -s https://<%- nvcfFunctionId %>.invocation.api.nvcf.nvidia.com/v1/audio/transcriptions \ -H "Authorization: Bearer <%- apiKey %>" \ -F language=en-GB \ -F file="@" ``` ## Inference using gRPC API Instructions below demonstrate usage of <%- name %> model using Python gRPC client. For full API details refer gRPC API Reference. ### Prerequisites You will need a system with Git and Python 3+ installed. ### Install Riva Python Client ```bash pip install -U nvidia-riva-client ``` ### Download Python Client Download Python client code by cloning Python Client Repository. ```bash git clone https://github.com/nvidia-riva/python-clients.git ``` ### Run Python Client Open a command terminal and execute below command to transcribe audio. Make sure you have a speech file in 16-bit Mono format in WAV/OGG/OPUS container. If you have generated the API key, it will be auto-populated in the command. The model auto-detects the spoken language across 25 supported European languages. Pass the appropriate BCP-47 language code via `--language-code` (e.g. `en-GB`, `de-DE`, `fr-FR`, `es-ES`). ```bash python python-clients/scripts/asr/transcribe_file_offline.py \ --server grpc.nvcf.nvidia.com:443 --use-ssl \ --metadata function-id "<%- nvcfFunctionId %>" \ --metadata "authorization" "Bearer <%- apiKey %>" \ --language-code en-GB \ --word-time-offsets --automatic-punctuation \ --input-file ``` ### Support for gRPC clients in other programming languages Riva uses gRPC APIs. Proto files can be downloaded from Riva gRPC Proto files and compiled to target language using Protoc compiler. Example Riva clients in C++ and Python languages are provided below. * Python Client Repository * C++ Client Repository 47:T864,## Step 1. Generate API Key ::generate-api-key ## Step 2. Pull and Run the NIM ```bash $ docker login nvcr.io Username: $oauthtoken Password: ``` Refer to [Supported Models](https://docs.nvidia.com/nim/speech/latest/reference/support-matrix/asr.html#asr-parakeet-06b-tdt) for full list of models. The Parakeet TDT v3 multilingual model is served from the `parakeet-0.6b-tdt` NIM by selecting the `multi` profile. ```bash export NGC_API_KEY= docker run -it --rm --name=parakeet-tdt \ --runtime=nvidia \ --gpus '"device=0"' \ --shm-size=8GB \ -e NGC_API_KEY \ -e NIM_HTTP_API_PORT=9000 \ -e NIM_GRPC_API_PORT=50051 \ -p 9000:9000 \ -p 50051:50051 \ -e NIM_TAGS_SELECTOR=type=multi \ nvcr.io/nim/nvidia/parakeet-0.6b-tdt:latest ``` It may take a up to 30 minutes depending on your network speed, for the container to be ready and start accepting requests from the time the docker container is started. ## Step 3. Test the NIM Open a new terminal and run following command to check if the service is ready to handle inference requests ```bash curl -X 'GET' 'http://localhost:9000/v1/health/ready' ``` If the service is ready, you get a response similar to the following. ```bash {"ready":true} ``` Install the Riva Python client package ```bash sudo apt-get install python3-pip pip install -U nvidia-riva-client ``` Download Riva sample clients ```bash git clone https://github.com/nvidia-riva/python-clients.git ``` Run Speech to Text inference in offline modes. Riva ASR supports Mono, 16-bit audio in WAV, OPUS and FLAC formats. The model auto-detects the spoken language across 25 supported European languages; pass the appropriate BCP-47 code via `--language-code` (e.g. `en-GB`, `de-DE`, `fr-FR`, `es-ES`). ```bash python3 python-clients/scripts/asr/transcribe_file_offline.py --server 0.0.0.0:50051 \ --language-code en-GB --word-time-offsets --automatic-punctuation \ --input-file ``` For more details on getting started with this NIM, visit the [Riva ASR NIM Docs](https://docs.nvidia.com/nim/speech/latest/asr/index.html). 3a:[["$","div",".3",{"className":"relative","children":[["$","div",null,{"className":"relative flex min-h-[194px] w-full flex-col overflow-hidden p-md sm:p-ml bg-manitoulinLightBlack","children":[["$","$L3d",null,{"alt":"","className":"object-cover object-center !left-1/2 z-10 !w-1/2 brightness-20 lg:brightness-75","fill":true,"priority":true,"src":"https://assets.ngc.nvidia.com/products/api-catalog/images/parakeet-tdt-0_6b.jpg","style":{"maskImage":"linear-gradient(to left, rgba(0,0,0,1), rgba(0,0,0,0))"}}],["$","div",null,{"className":"z-30 mx-auto flex w-full max-w-screen-xl min-w-0 flex-1 flex-col items-start justify-between gap-md lg:flex-row lg:gap-ml","children":["$","section",null,{"aria-describedby":"resource-description","aria-labelledby":"resource-name","className":"flex w-full min-w-0 flex-col lg:flex-1","children":["$","div",null,{"className":"flex flex-col gap-1.5 whitespace-pre-wrap","children":[["$","$L3e",null,{"className":"w-fit text-ml text-n300 lowercase hover:text-n050","href":"/nvidia","children":"NVIDIA"}],["$","div",null,{"className":"flex flex-col items-start gap-3 lg:flex-row lg:items-center","children":[["$","h1",null,{"className":"max-w-full text-lg leading-heading font-medium break-words text-primary","data-testid":"resource-name","id":"resource-name","children":"parakeet-tdt-0.6b"}],[["$","span","AVAILABLE",{"className":"nv-badge nv-badge--kind-solid h-[18px] py-0","data-testid":"nv-badge","ref":"$undefined","children":["$undefined","Downloadable"]}]],[["$","button","#1 on Throughput HF ASR Leaderboard",{"className":"nv-tag nv-tag--color-yellow nv-density-standard nv-tag--kind-outline [[readonly]]:pointer-events-none py-0 align-middle leading-body h-[calc(var(--spacing-density-sm)*2+var(--text-12)+var(--border-width-1)*2)] flex-row","data-readonly":"$undefined","data-selected":"$undefined","data-testid":"nv-tag-root","ref":"$undefined","type":"button","children":[["$","$L3f",null,{"children":[["$","svg","award-trophy:fill",{"data-src":"https://brand-assets.cne.ngc.nvidia.com/assets/icons/3.8.0/fill/award-trophy.svg","height":"1em","width":"1em","display":"inline-block","data-icon-name":"award-trophy","data-cache":"disabled","className":"text-[length:inherit]","ref":"$undefined"}]]}],["$","$L3e",null,{"href":"https://huggingface.co/spaces/hf-audio/open_asr_leaderboard","target":"_blank","children":["#",1," on ","Throughput HF ASR Leaderboard"]}]]}]]]}],["$","p",null,{"className":"max-w-[85ch] text-ms text-primary","id":"resource-description","children":"Multilingual ASR across 25 European languages with punctuation, capitalization, and word timestamps"}],["$","div",null,{"className":"flex w-full flex-col items-start gap-md lg:flex-row lg:items-end lg:justify-between","children":[["$","div",null,{"className":"flex w-full min-w-0 flex-wrap gap-sm lg:flex-1","children":["$","$L40",null,{"artifactType":"ENDPOINT","density":"standard","labels":["ASR","Multilingual","NVIDIA NIM","NVIDIA Riva","nimType:endpoint:nim_type_run_anywhere","speech-to-text"],"searchRoutePath":"/models"}]}],["$","div",null,{"className":"z-30 flex w-full flex-row justify-start gap-ms sm:justify-end lg:w-auto lg:shrink-0","children":["$undefined",["$","$L41",null,{"cta":{"text":"Run Anywhere - Notify Me","url":"https://www.nvidia.com/en-us/ai/nim-notifyme/","nim_available_override_url":"https://catalog.ngc.nvidia.com/orgs/nim/teams/nvidia/containers/parakeet-tdt-0.6b-v3"},"isNIMAccessRestricted":"$undefined","isNIMAvailable":true}]]}]]}]]}]}]}],["$","div",null,{"className":"pointer-events-none absolute right-0 bottom-0 left-0 z-20 h-30 bg-gradient-to-t to-transparent from-manitoulinLightBlack via-manitoulinLightBlack/70"}]]}],["$","div",null,{"className":"px-4 mx-auto flex max-w-screen-xl flex-col pt-0 lg:px-0","data-body-background":"light-black","children":["$","section",null,{"className":"flex w-full flex-col justify-between rounded-4 border-manitoulinBorderColor bg-manitoulinDarkBlack sm:rounded-6 lg:rounded-8 p-ms lg:p-lg","children":["$","$L42",null,{"artifact":{"artifactType":"ENDPOINT","name":"parakeet-tdt-0_6b","displayName":"parakeet-tdt-0.6b","publisher":"nvidia","shortDescription":"Multilingual ASR across 25 European languages with punctuation, capitalization, and word timestamps","logo":"https://assets.ngc.nvidia.com/products/api-catalog/images/parakeet-tdt-0_6b.jpg","labels":"$3a:0:props:children:0:props:children:1:props:children:props:children:props:children:3:props:children:0:props:children:props:labels","attributes":[{"key":"AVAILABLE","value":"true"},{"key":"PREVIEW","value":"false"}],"updatedDate":"2026-09-08T07:26:44.914Z","bias":"Field | Response\n---------------------------------------------------------------------------------------------------|---------------\nParticipation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing | None\nMeasures taken to mitigate against unwanted bias | None","canGuestDownload":true,"createdDate":"2026-09-08T07:26:44.914Z","description":"$43","explainability":"$44","isPublic":true,"isReadOnly":true,"orgName":"qc69jvmznzxy","privacy":"$45","safetyAndSecurity":"Field | Response\n---------------------------------------------------|----------------------------------\nModel Application(s) | Speech to Text Transcription\nDescribe the life critical impact | None\nUse Case Restrictions | Abide by [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/legalcode.en) License\nModel and dataset restrictions | The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training, and dataset license constraints adhered to."},"endpointName":"parakeet-tdt-0_6b","endpointSpec":{"namespace":"qc69jvmznzxy","nvcfFunctionId":"None","createdDate":"2026-09-08T07:26:45.251Z","attributes":{"apiDocsUrl":"https://docs.nvidia.com/nim/speech/latest/reference/api-references/asr/protos.html","termsOfUse":"GOVERNING TERMS: Your use of this API is governed by the NVIDIA API Trial Terms of Service and the use of this model is governed by the NVIDIA Open Model License Agreement. \n","showUnavailableBanner":false,"cta":"$3a:0:props:children:0:props:children:1:props:children:props:children:props:children:3:props:children:1:props:children:1:props:cta","usage":"$46","deploy":[{"label":"Linux with Docker","filename":"linux.md","contents":"$47"}],"leaderboard":[{"name":"Throughput HF ASR Leaderboard","url":"https://huggingface.co/spaces/hf-audio/open_asr_leaderboard","position":1}],"modelCapability":{"functionCalling":false}},"artifactName":"parakeet-tdt-0_6b"},"isLoginRequired":true,"playgroundElement":"$48","publisherName":"nvidia","searchParams":{"model":"parakeet-tdt-0_6b"},"tabs":[{"label":"Experience","bare":true,"icon":"play","kind":"default","id":"experience","href":"/nvidia/parakeet-tdt-0_6b","aliases":["/nvidia/parakeet-tdt-0_6b/experience"]},{"label":"Model Card","icon":"book","kind":"default","id":"modelcard","href":"/nvidia/parakeet-tdt-0_6b/modelcard","aliases":"$undefined"},{"label":"Try API","kind":"default","id":"api","href":"/nvidia/parakeet-tdt-0_6b/api","aliases":"$undefined"},{"label":"Deploy","kind":"default","id":"deploy","href":"/nvidia/parakeet-tdt-0_6b/deploy","aliases":"$undefined"}]}]}]}]]}]] 3b:["$","$L49",null,{"spec":"$3a:0:props:children:1:props:children:props:children:props:endpointSpec"}]