{"schema":"catalog-search/v2","items":[{"kind":"record","record":{"id":"src_01M569EH6TMT2XNNCMYNRZ573R","type":"source","title":"faster-whisper","summary":"The README describes faster-whisper as a reimplementation of OpenAI's Whisper model using the CTranslate2 inference engine, for speech transcription.\nIt claims up to 4 times the speed of openai/whisper at the same accuracy with less memory.","snippet":"README describes <b>faster</b>-<b>whisper</b> as a reimplementation of OpenAI's <b>Whisper</b> model using the CTranslate2 inference engine, for speech transcription.\nIt claims up to 4 <b>times</b>","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M569EH6TMT2XNNCMYNRZ573R","api_url":"https://veyrul.com/v1/records/src_01M569EH6TMT2XNNCMYNRZ573R","updated_at":"2026-11-07T11:17:44.219Z","score":8.4923}},{"kind":"record","record":{"id":"src_01M569EGKB12MJ1XPDKRGFZY25","type":"source","title":"WhisperX","summary":"The README says the repository provides fast automatic speech recognition (70x realtime with large-v2) with word-level timestamps and speaker diarization.","snippet":"<b>Whisper</b> pipeline adding batched inference (README: 70x realtime with large-v2 via <b>faster</b>-<b>whisper</b>), wav2vec2 forced alignment for word-level timestamps, and pyannote speaker diarization. Practical","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M569EGKB12MJ1XPDKRGFZY25","api_url":"https://veyrul.com/v1/records/src_01M569EGKB12MJ1XPDKRGFZY25","updated_at":"2026-11-07T13:36:28.201Z","score":1.1773}},{"kind":"record","record":{"id":"src_01M6SAX7KG4XF3N22DW1KY4B53","type":"source","title":"pycaps video caption tool","summary":"A Python library and CLI that renders animated CSS-styled subtitles onto videos.","snippet":"install from GitHub; README recommends the [all] extra, [base] adds <b>whisper</b> and playwright, [base,<b>faster</b>-<b>whisper</b>] adds a <b>faster</b> backend (--<b>whisper</b>-backend <b>faster</b>), [fast","status":"active","categories":["T10"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6SAX7KG4XF3N22DW1KY4B53","api_url":"https://veyrul.com/v1/records/src_01M6SAX7KG4XF3N22DW1KY4B53","updated_at":"2026-11-06T21:05:50.660Z","score":1.1291}},{"kind":"record","record":{"id":"src_01M6SRZHPDDDXCWHQMBQ8GSJRD","type":"source","title":"whisper.cpp","summary":"Dependency-free C/C++ port of OpenAI Whisper speech recognition with CPU, Metal, CUDA, Vulkan and other backends and integer quantization.","snippet":"Metal, CUDA, Vulkan and other backends and integer quantization.\n\nUseful for: spoken <b>audio</b>: <b>transcribe</b> interviews, podcasts or voiceover to <b>text</b> and subtitles offline\n\nFirst action: DOCUMENTED (README Quick","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6SRZHPDDDXCWHQMBQ8GSJRD","api_url":"https://veyrul.com/v1/records/src_01M6SRZHPDDDXCWHQMBQ8GSJRD","updated_at":"2026-11-07T01:11:46.602Z","score":0.1307}},{"kind":"record","record":{"id":"src_01M6S6W9MAPP8Y9T04S82VPRKD","type":"source","title":"Voice-Pro","summary":"Gradio web UI combining YouTube download, voice separation, speech recognition, translation, TTS and dubbing.","snippet":"download, voice separation, speech recognition, translation, TTS and dubbing.\n\nUseful for: spoken <b>audio</b>: <b>transcribe</b>, translate and dub video <b>audio</b>\n\nFirst action: DOCUMENTED (README Installation, current v4 section","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6S6W9MAPP8Y9T04S82VPRKD","api_url":"https://veyrul.com/v1/records/src_01M6S6W9MAPP8Y9T04S82VPRKD","updated_at":"2026-11-06T19:55:25.669Z","score":0.1133}},{"kind":"record","record":{"id":"src_01M569EGSVM7N346FE5S54X20Z","type":"source","title":"OpenAI Whisper","summary":null,"snippet":"General-purpose multilingual speech recognition, speech translation, language ID and","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M569EGSVM7N346FE5S54X20Z","api_url":"https://veyrul.com/v1/records/src_01M569EGSVM7N346FE5S54X20Z","updated_at":"2026-10-18T01:24:31.262Z","score":0.0506}},{"kind":"record","record":{"id":"src_01M6VAXAV9AS4FJSR8DJ3HTD0F","type":"source","title":"SpeechBrain: PyTorch toolkit for speech and text processing, README overview","summary":"The README presents SpeechBrain as a PyTorch toolkit for speech and text processing in conversational AI.","snippet":"PyTorch toolkit for speech and <b>text</b> processing in conversational AI.\n\nUseful for: - <b>Transcribe</b> speech with a pretrained model in a few lines.\n- Train or adapt models from the shared","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6VAXAV9AS4FJSR8DJ3HTD0F","api_url":"https://veyrul.com/v1/records/src_01M6VAXAV9AS4FJSR8DJ3HTD0F","updated_at":"2026-11-07T15:44:22.854Z","score":0.0149}},{"kind":"record","record":{"id":"src_01M6VD2QGZQG09PCF0RNRNDPH2","type":"source","title":"ESPnet: end-to-end speech processing toolkit, README overview","summary":"The README presents ESPnet as a PyTorch toolkit for many speech tasks.","snippet":"README presents ESPnet as a PyTorch toolkit for many speech tasks.\n\nUseful for: - <b>Transcribe</b> <b>audio</b> with a pretrained model from the command line.\n- Train or reproduce models with","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6VD2QGZQG09PCF0RNRNDPH2","api_url":"https://veyrul.com/v1/records/src_01M6VD2QGZQG09PCF0RNRNDPH2","updated_at":"2026-11-07T16:22:16.820Z","score":0.0137}},{"kind":"record","record":{"id":"src_01M6V5KVEB336053WGZ1R7RW55","type":"source","title":"Vosk: offline speech recognition toolkit, README overview","summary":"The README presents Vosk as an offline open source speech recognition toolkit for 20+ languages and dialects.","snippet":"offline speech recognition to chatbots, smart home devices or virtual assistants.\n- Create subtitles or <b>transcribe</b> lectures and interviews.\n- Run on devices from a Raspberry Pi or phone up to large","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6V5KVEB336053WGZ1R7RW55","api_url":"https://veyrul.com/v1/records/src_01M6V5KVEB336053WGZ1R7RW55","updated_at":"2026-11-07T14:11:49.291Z","score":0.0135}},{"kind":"record","record":{"id":"src_01M6V8ETC68RASAA4MDB69DDMQ","type":"source","title":"FunASR: speech recognition toolkit for offline, streaming and edge use, README overview","summary":"The README presents FunASR as a speech recognition toolkit for offline, streaming and edge use, with models for ASR, VAD, punctuation, speaker work and more.","snippet":"edge use, with models for ASR, VAD, punctuation, speaker work and more.\n\nUseful for: - <b>Transcribe</b> <b>audio</b> with Fun-ASR-Nano through the Transformers CPU quickstart or the toolkit.\n- Choose","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6V8ETC68RASAA4MDB69DDMQ","api_url":"https://veyrul.com/v1/records/src_01M6V8ETC68RASAA4MDB69DDMQ","updated_at":"2026-11-07T15:01:30.080Z","score":0.0132}},{"kind":"record","record":{"id":"src_01M6V4XH72EDN921JZP1HHD1NR","type":"source","title":"llamafile: run LLMs from a single-file executable, README overview","summary":"The README presents llamafile as a way to distribute and run LLMs as one file that runs locally with no installation, using llama.cpp and Cosmopolitan Libc.","snippet":"Hand someone a single executable that runs a local model without an installer.\n- <b>Transcribe</b> or translate <b>audio</b> files with the bundled whisperfile tool.\n- Run the llamafile binary with your","status":"active","categories":["T09"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6V4XH72EDN921JZP1HHD1NR","api_url":"https://veyrul.com/v1/records/src_01M6V4XH72EDN921JZP1HHD1NR","updated_at":"2026-11-07T13:59:37.920Z","score":0.013}},{"kind":"record","record":{"id":"src_01M6FM34K5BGVY3SZQT772YGP5","type":"source","title":"OpenMOSS-Team/MOSS-Transcribe-Diarize","summary":"Hugging Face model OpenMOSS-Team/MOSS-Transcribe-Diarize at revision 704aa4a9c304: audio-text-to-text; library transformers; declared licence apache-2.0; not gated. Metadata-only index entry; not tested guidance.","snippet":"verified): papers arxiv:2601.01554\ndeclared tags (declared by the uploader, not verified): moss_<b>transcribe</b>_diarize, text2text-generation, moss, <b>audio</b>, speech, asr, diarization, timestamp-asr, long-form-<b>audio</b>, multimodal","status":"active","categories":["T09"],"evidence":{"inspection":"metadata_checked","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6FM34K5BGVY3SZQT772YGP5","api_url":"https://veyrul.com/v1/records/src_01M6FM34K5BGVY3SZQT772YGP5","updated_at":"2026-11-03T02:33:57.057Z","score":0.012}},{"kind":"record","record":{"id":"src_01M6SDPJNT32P625PXV8EVMPDD","type":"source","title":"Music To MIDI","summary":"Local-first audio-to-MIDI converter with CLI batch mode for multi-track transcription, optional lyrics embedding and SoundFont rendering.","snippet":"mode for multi-track transcription, optional lyrics embedding and SoundFont rendering.\n\nUseful for: music: <b>transcribe</b> songs into multi-track MIDI\n\nFirst action: DOCUMENTED (README, Chinese): choose one route. Source route","status":"active","categories":["T14"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6SDPJNT32P625PXV8EVMPDD","api_url":"https://veyrul.com/v1/records/src_01M6SDPJNT32P625PXV8EVMPDD","updated_at":"2026-11-06T21:54:38.348Z","score":0.0119}},{"kind":"record","record":{"id":"src_01M6S7ETYP0WR4A31327B1CET0","type":"source","title":"Parakeet TDT 0.6B v2 - fast English ASR with timestamps","summary":"600M-parameter English ASR with punctuation, capitalisation and word/segment/char timestamps; up to ~24 min per pass; RTFx 3380 at batch 128 on the HF Open ASR leaderboard; WER tables incl. noise and 8 kHz telephony.","snippet":"First action: pip install -U nemo_toolkit[\"asr\"], load nvidia/parakeet-tdt-0.6b-v2 with ASRModel.from_pretrained, then <b>transcribe</b> a 16 kHz mono wav with timestamps=True.\n\nPrerequisites and limits: Governed","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6S7ETYP0WR4A31327B1CET0","api_url":"https://veyrul.com/v1/records/src_01M6S7ETYP0WR4A31327B1CET0","updated_at":"2026-11-06T20:05:33.227Z","score":0.0114}},{"kind":"record","record":{"id":"dsc_01M569J6GWPKEHV00378AVC4XW","type":"discussion","title":"Does neural denoising (DeepFilterNet-class) hurt or help downstream ASR and ADR matching?","summary":null,"snippet":"that degrade ASR or voice cloning. What order of operations (denoise before or after <b>Whisper</b>, separation before cloning) maximises transcript accuracy and voice fidelity on location sound?\n\nWhy this","status":"open","categories":["T13","T14"],"evidence":{"inspection":"discovered","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/discussions/dsc_01M569J6GWPKEHV00378AVC4XW","api_url":"https://veyrul.com/v1/records/dsc_01M569J6GWPKEHV00378AVC4XW","updated_at":"2026-10-18T01:21:04.531Z","score":0.0091}},{"kind":"record","record":{"id":"src_01M67C6VJQHTEB7RQHCS26KESJ","type":"source","title":"Sunbird/asr-whisper-large-v2-salt","summary":"Hugging Face model Sunbird/asr-whisper-large-v2-salt at revision 54f8e3f92895: automatic-speech-recognition; library transformers; declared licence not declared; base openai/whisper-large-v2; gated (auto); finetune variant. Metadata-only index entry; not tested guidance.","snippet":"trained or evaluated on datasets Sunbird/salt\ndeclared tags (declared by the uploader, not verified): <b>whisper</b>, lg, en, nyn, ach, teo, lgg\nbase models: [\"openai/whisper-large-v2\"]\ncreated","status":"active","categories":["T13"],"evidence":{"inspection":"metadata_checked","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M67C6VJQHTEB7RQHCS26KESJ","api_url":"https://veyrul.com/v1/records/src_01M67C6VJQHTEB7RQHCS26KESJ","updated_at":"2026-10-30T21:42:14.829Z","score":0.0076}},{"kind":"record","record":{"id":"src_01M6EFB4HVAVHPWYW918RQ6SG2","type":"source","title":"cstr/ark-asr-3b-GGUF","summary":"Hugging Face model cstr/ark-asr-3b-GGUF at revision 3f228f0d7835: automatic-speech-recognition; library gguf; declared licence other; base AutoArk-AI/ARK-ASR-3B; not gated; quantized variant. Metadata-only index entry; not tested guidance.","snippet":"quantized from AutoArk-AI/ARK-ASR-3B\ndeclared tags (declared by the uploader, not verified): asr, <b>whisper</b>, qwen2, crispasr, ggml","status":"active","categories":["T13"],"evidence":{"inspection":"metadata_checked","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6EFB4HVAVHPWYW918RQ6SG2","api_url":"https://veyrul.com/v1/records/src_01M6EFB4HVAVHPWYW918RQ6SG2","updated_at":"2026-11-02T15:51:41.796Z","score":0.0074}},{"kind":"record","record":{"id":"src_01M569EGNJWQ9KRD31AHZVQS3S","type":"source","title":"NVIDIA Parakeet TDT 0.6B v3","summary":"Hugging Face model nvidia/parakeet-tdt-0.6b-v3 at revision 541d1f99c6b0: automatic-speech-recognition; library transformers; declared licence cc-by-4.0; not gated. Metadata-only index entry; not tested guidance.","snippet":"declared tags (declared by the uploader, not verified): nemo, parakeet_tdt, feature-extraction, speech, <b>audio</b>, Transducer, Transformer, TDT, FastConformer, Conformer, pytorch, hf-asr-leaderboard","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"reported_by_source","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M569EGNJWQ9KRD31AHZVQS3S","api_url":"https://veyrul.com/v1/records/src_01M569EGNJWQ9KRD31AHZVQS3S","updated_at":"2026-11-01T17:13:17.760Z","score":0.007}},{"kind":"record","record":{"id":"src_01M6JW444RCANHB9JMZ5H3ER7N","type":"source","title":"WJ88/NVIDIA-Parakeet-TDT-0.6B-v2-INT8-Real-Time-Mic-Transcription","summary":"Hugging Face Space WJ88/NVIDIA-Parakeet-TDT-0.6B-v2-INT8-Real-Time-Mic-Transcription at revision d937c776cb3e: SDK gradio; declared licence not declared; access gating not reported. Metadata-only index entry; not tested guidance.","snippet":"declared tags (declared by the uploader, not verified): <b>audio</b>, speech-recognition, asr, real-<b>time</b>, cpu, nvidia, parakeet, microphone, voice, speech, browser, nemo, huggingface\ncreated","status":"active","categories":["T09"],"evidence":{"inspection":"metadata_checked","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M6JW444RCANHB9JMZ5H3ER7N","api_url":"https://veyrul.com/v1/records/src_01M6JW444RCANHB9JMZ5H3ER7N","updated_at":"2026-11-04T08:52:01.270Z","score":0.0002}},{"kind":"record","record":{"id":"src_01M569EGVGE98442AG0PZ67B35","type":"source","title":"pyannote.audio (speaker diarization toolkit)","summary":"The README says pyannote.audio is an open-source toolkit written in Python for speaker diarization, based on PyTorch, with pretrained models and pipelines.","snippet":"pipelines.\n\nUseful for: - Split a recording into speaker turns with start and stop <b>times</b> using the community-1 open-source pipeline run locally.\n- Use the premium precision-2 service through","status":"active","categories":["T13"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed"},"url":"https://veyrul.com/sources/src_01M569EGVGE98442AG0PZ67B35","api_url":"https://veyrul.com/v1/records/src_01M569EGVGE98442AG0PZ67B35","updated_at":"2026-11-07T13:39:46.096Z","score":0.0001}},{"kind":"guide_preview","preview":{"schema":"guide-preview/v1","id":"art_01M6TXAJ5VZE8A0804QMWFVYXV","type":"article","revision":1,"title":"Transcribe audio to timed text with faster-whisper","coverage":"For editors and podcasters who want a timestamped transcript of one audio file from a Python script. You end with a short script that writes start and end times for each segment. It is based on the project README and was not run.","example":"A hypothetical script transcribes interview.mp3 on CPU with the silence filter on and writes timed lines to a text file; the example was not run.","tips":["The transcription result is a generator, so nothing happens until you loop over it or wrap it in a list.","Compare the transcript with the audio at the start, middle and end rather than trusting the first lines.","When comparing against another Whisper tool, match the beam size and CPU thread count or the speed comparison is unfair."],"categories":["T13","T10"],"tags":["faster-whisper","transcription","whisper","ctranslate2","speech-to-text","vad","local-audio"],"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed","review_revision":null},"media":[]}}],"truncated":false,"has_more":true,"next_cursor":"eyJyZWZlcmVuY2UiOiJNakEiLCJwcmV2aWV3IjpudWxsLCJxIjoiVHJhbnNjcmliZSBhdWRpbyB0byB0aW1lZCB0ZXh0IHdpdGggZmFzdGVyLXdoaXNwZXIiLCJ0eXBlcyI6W10sImNhdGVnb3JpZXMiOltdfQ","ordering":"bounded_reference_and_preview_lanes"}