据 AlphaMoat 统计,截至 2026-10,Faster Whisper 累计下载 16,749 次、获 11 个星标,按累计下载量在本次参与排名的 10,914 个 AI 技能中排名第 1,685 位,在「设计创意」分类(1,116 个)中排名第 229 位。
列表默认按累计下载量排序,也可按星标排序;详情页排名按累计下载量计算。
| 平台 | 累计下载 | 星标 | 评论 |
|---|---|---|---|
| ClawHub ↗ | 8,948 | 5 | - |
| SkillHub ↗ | 1.7万 | 11 | — |
各平台的计数口径不同,统计时间不一致。
基于 faster-whisper(Whisper 的 CTranslate2 实现)在本地转录,可自动检测语言、批量处理、估算进度,并支持搜索转录内容和章节检测。可选说话人分离 --diarize(需要 pyannote.audio),URL/YouTube 输入使用 yt-dlp。默认最快用法是 ./scripts/transcribe audio.mp3;只有用户明确要求时才增加额外标志。默认模型为 distil-large-v3.5;基础转录通过 PyAV 解码,不要求 ffmpeg。
资料发布于 2026-10-27
以下是根据技能文档生成的示例,不代表实际运行结果。
2026-09 起统计方法调整:同名技能按作者分别统计,并覆盖全部技能。图中虚线为调整前的月份,与之后的数据不直接比较,也不计算跨越调整的增长。
| 月份 | 累计下载 | 环比 | 星标 |
|---|---|---|---|
| 2026-10 | 1.7万 | — | 11 |
| 2026-09 | 9,440 | — | 11 |
| 2026-08 | 8,464 | +3.6% | 11 |
| 2026-07 | 8,167 | +6.7% | 5 |
| 2026-06 | 7,654 | +4.3% | 10 |
| 2026-05 | 7,338 | — | 9 |
以下是作者发布的原始文件,按纯文本显示,本站未做修改或审核;其中的链接不可点击,命令请核实后再运行。抓取于 2026-10-23。
---
name: faster-whisper
description: "Local speech-to-text using faster-whisper. 4-6x faster than OpenAI Whisper with identical accuracy; GPU acceleration enables ~20x realtime transcription. SRT/VTT/TTML/CSV subtitles, speaker diarization, URL/YouTube input, batch processing with ETA, transcript search, chapter detection, per-file language map."
version: 1.5.1
author: ThePlasmak
homepage: https://github.com/ThePlasmak/faster-whisper
tags:
[
"audio",
"transcription", "whisper",
"speech-to-text",
"ml",
"cuda",
"gpu",
"subtitles",
"diarization",
"podcast",
"chapters",
"search",
"csv",
"ttml",
"batch",
]
platforms: ["linux", "macos", "wsl2"]
metadata:
{
"openclaw":
{
"emoji": "🗣️",
"requires":
{
"bins": ["python3"],
"optionalBins": ["ffmpeg", "yt-dlp"],
"optionalPaths": ["~/.cache/huggingface/token"],
},
},
}
---
# Faster Whisper
Local speech-to-text using faster-whisper — a CTranslate2 reimplementation of OpenAI's Whisper that runs **4-6x faster** with identical accuracy. With GPU acceleration, expect **~20x realtime** transcription (a 10-minute audio file in ~30 seconds).
## When to Use
Use this skill when you need to:
- **Transcribe audio/video files** — meetings, interviews, podcasts, lectures, YouTube videos
- **Generate subtitles** — SRT, VTT, ASS, LRC, or TTML broadcast-standard subtitles
- **Identify speakers** — diarization labels who said what (`--diarize`)
- **Transcribe from URLs** — YouTube links and direct audio URLs (auto-downloads via yt-dlp)
- **Transcribe podcast feeds** — `--rss <feed-url>` fetches and transcribes episodes
- **Batch process files** — glob patterns, directories, skip-existing support; ETA shown automatically
- **Convert speech to text locally** — no API costs, works offline (after model download)
- **Translate to English** — translate any language to English with `--translate`
- **Do multilingual transcription** — supports 99+ languages with auto-detection
- **Transcribe a batch of files in different languages** — `--language-map` assigns a different language per file
- **Transcribe multilingual audio** — `--multilingual` for mixed-language audio
- **Transcribe audio with specific terms** — use `--initial-prompt` for jargon-heavy content or any other terms to look out for
- **Preprocess noisy audio (before transcription)** — `--normalize` and `--denoise` before transcription
- **Stream output** — `--stream` shows segments as they're transcribed
- **Clip time ranges** — `--clip-timestamps` to transcribe specific sections
- **Search the transcript** — `--search "term"` finds all timestamps where a word/phrase appears
- **Detect chapters** — `--detect-chapters` finds section breaks from silence gaps
- **Export speaker audio** — `--export-speakers DIR` saves each speaker's turns as separate WAV files
- **Spreadsheet output** — `--format csv` produces a properly-quoted CSV with timestamps
**Trigger phrases:**
"transcribe this audio", "convert speech to text", "what did they say", "make a transcript",
"audio to text", "subtitle this video", "who's speaking", "translate this audio", "translate to English",
"find where X is mentioned", "search transcript for", "when did they say", "at what timestamp",
"add chapters", "detect chapters", "find breaks in the audio", "table of contents for this recording",
"TTML subtitles", "DFXP subtitles", "broadcast format subtitles", "Netflix format",
"ASS subtitles", "aegisub format", "advanced substation alpha", "mpv subtitles",
"LRC subtitles", "timed lyrics", "karaoke subtitles", "music player lyrics",
"HTML transcript", "confidence-colored transcript", "color-coded transcript",
"separate audio per speaker", "export speaker audio", "split by speaker",
"transcript as CSV", "spreadsheet output", "transcribe podcast", "podcast RSS feed",
"different languages in batch", "per-file language",
"transcribe in multiple formats", "srt and txt at the same time", "output both srt and text",
"remove filler words", "clean up ums and uhs", "strip hesitation sounds", "remove you know and I mean",
"transcribe left channel", "transcribe right channel", "stereo channel", "left track only",
"wrap subtitle lines", "character limit per line", "max chars per subtitle",
"detect paragraphs", "paragraph breaks", "group into paragraphs", "add paragraph spacing"
**⚠️ Agent guidance — keep invocations minimal:**
_CORE RULE: default command (`./scripts/transcribe audio.mp3`) is the fastest path — add flags only when the user explicitly asks for that capability._
**Transcription:**
- Only add `--diarize` if the user asks "who said what" / "identify speakers" / "label speakers"
- Only add `--format srt/vtt/ass/lrc/ttml` if the user asks for subtitles/captions in that format
- Only add `--format csv` if the user asks for CSV or spreadsheet output
- Only add `--word-timestamps` if the user needs word-level timing
- Only add `--initial-prompt` if there's domain-specific jargon to prime
- Only add `--translate` if the user wants non-English audio translated to English
- Only add `--normalize`/`--denoise` if the user mentions bad audio quality or noise
- Only add `--stream` if the user wants live/progressive output for long files
- Only add `--clip-timestamps` if the user wants a specific time range
- Only add `--temperature 0.0` if the model is hallucinating on music/silence
- Only add `--vad-threshold` if VAD is aggressively cutting speech or including noise
- Only add `--min-speakers`/`--max-speakers` when you know the speaker count
- Only add `--hf-token` if the token is not cached at `~/.cache/huggingface/token`
- Only add `--max-words-per-line` for subtitle readability on long segments
- Only add `--filter-hallucinations` if the transcript contains obvious artifacts (music markers, duplicates)
- Only add `--merge-sentences` if the user asks for sentence-level subtitle cues
- Only add `--clean-filler` if the user asks to remove filler words (um, uh, you know, I mean, hesitation sounds)
- Only add `--channel left|right` if the user mentions stereo tracks, dual-channel recordings, or asks for a specific channel
- Only add `--max-chars-per-line N` when the user specifies a character limit per subtitle line (e.g., "Netflix format", "42 chars per line"); takes priority over `--max-words-per-line`
- Only add `--detect-paragraphs` if the user asks for paragraph breaks or structured text output; `--paragraph-gap` (default 3.0s) only if they want a custom gap
- Only add `--speaker-names "Alice,Bob"` when the user provides real names to replace SPEAKER_1/2 — always requires `--diarize`
- Only add `--hotwords WORDS` when the user names specific rare terms not well served by `--initial-prompt`; prefer `--initial-prompt` for general domain jargon
- Only add `--prefix TEXT` when the user knows the exact words the audio starts with
- Only add `--detect-language-only` when the user only wants to identify the language, not transcribe
- Only add `--stats-file PATH` if the user asks for performance stats, RTF, or benchmark info
- Only add `--parallel N` for large CPU batch jobs; GPU handles one file efficiently on its own — don't add for single files or small batches
- Only add `--retries N` for unreliable inputs (URLs, network files) where transient failures are expected
- Only add `--burn-in OUTPUT` only when user explicitly asks to embed/burn subtitles into the video; requires ffmpeg and a video file input
- Only add `--keep-temp` when the user may re-process the same URL to avoid re-downloading
- Only add `--output-template` when user specifies a custom naming pattern in batch mode
- **Multi-format output** (`--format srt,text`): only when user explicitly wants multiple formats in one pass; always pair with `-o <dir>`
- Any word-level feature auto-runs wav2vec2 alignment (~5-10s overhead)
- `--diarize` adds ~20-30s on top of that
**Search:**
- Only add `--search "term"` when the user asks to find/locate/search for a specific word or phrase in audio
- `--search` **replaces** the normal transcript output — it prints only matching segments with timestamps
- Add `--search-fuzzy` only when the user mentions approximate/partial matching or typos
- To save search results to a file, use `-o results.txt`
**Chapter detection:**
- Only add `--detect-chapters` when the user asks for chapters, sections, a table of contents, or "where does the topic change"
- Default `--chapter-gap 8` (8-second silence = new chapter) works for most podcasts/lectures; tune down for dense content
- `--chapter-format youtube` (default) outputs YouTube-ready timestamps; use `json` for programmatic use
- **Always use `--chapters-file PATH`** when combining chapters with a transcript output — avoids mixing chapter markers into the transcript text
- If the user only wants chapters (not the transcript), pipe stdout to a file with `-o /dev/null` and use `--chapters-file`
- **Batch mode limitation:** `--chapters-file` takes a single path — in batch mode, each file's chapters overwrite the previous. For batch chapter detection, omit `--chapters-file` (chapters print to stdout under `=== CHAPTERS (N) ===`) or use a separate run per file
**Speaker audio export:**
- Only add `--export-speakers DIR` when the user explicitly asks to save each speaker's audio separately
- Always pair with `--diarize` — it silently skips if no speaker labels are present
- Requires ffmpeg; outputs `SPEAKER_1.wav`, `SPEAKER_2.wav`, etc. (or real names if `--speaker-names` is set)
**Language map:**
- Only add `--language-map` in batch mode when the user has confirmed different languages across files
- Inline format: `"interview*.mp3=en,lecture*.mp3=fr"` — fnmatch globs on filename
- JSON file format: `@/path/to/map.json` where the file is `{"pattern": "lang_code"}`
**RSS / Podcast:**
- Only add `--rss URL` when the user provides a podcast RSS feed URL
- Default fetches 5 newest episodes; `--rss-latest 0` for all; `--skip-existing` to resume safely
- **Always use `-o <dir>`** with `--rss` — without it, all episode transcripts print to stdout concatenated, which is hard to use; each episode gets its own file when `-o <dir>` is set
**Output format for agent relay:**
- **Search results** (`--search`) → print directly to user; output is human-readable
- **Chapter output** → if no `--chapters-file`, chapters appear in stdout under `=== CHAPTERS (N) ===` header after the transcript; with `--format json`, chapters are also embedded in the JSON under `"chapters"` key
- **Subtitle formats** (SRT, VTT, ASS, LRC, TTML) → always write to `-o` file; tell the user the output path, never paste raw subtitle content
- **Data formats** (CSV, HTML, TTML, JSON) → always write to `-o` file; tell the user the output path, don't paste raw XML/CSV/HTML
- **ASS format** → for Aegisub, VLC, mpv; write to file and tell user they can open it in Aegisub or play it in VLC/mpv
- **LRC format** → timed lyrics for music players (Foobar2000, AIMP, VLC); write to file
- **Multi-format** (`--format srt,text`) → requires `-o <dir>`; each format goes to a separate file; tell user all paths written
- **JSON format** → useful for programmatic post-processing; not ideal to paste in full to user
- **Text/transcript** → safe to show directly to user for short files; summarise for long ones
- **Stats output** (`--stats-file`) → summarise key fields (duration, processing time, RTF) for the user rather than pasting raw JSON
- **Language detection** (`--detect-language-only`) → print the result directly; it's a single line
- **ETA** is printed automatically to stderr for batch jobs; no action needed
**When NOT to use:**
- Cloud-only environments without local compute
- Files <10 seconds where API call latency doesn't matter
**faster-whisper vs whisperx:**
This skill covers everything whisperx does — diarization (`--diarize`), word-level timestamps (`--word-timestamps`), SRT/VTT subtitles — so whisperx is not needed. Use whisperx only if you specifically need its pyannote pipeline or batch-GPU features not covered here.
## Quick Reference
| Task | Command | Notes |
| ------------------------------ | -------------------------------------------------------------------------------------- | --------------------------------------------------- |
| **Basic transcription** | `./scripts/transcribe audio.mp3` | Batched inference, VAD on, distil-large-v3.5 |
| **SRT subtitles** | `./scripts/transcribe audio.mp3 --format srt -o subs.srt` | Word timestamps auto-enabled |
| **VTT subtitles** | `./scripts/transcribe audio.mp3 --format vtt -o subs.vtt` | WebVTT format |
| **Word timestamps** | `./scripts/transcribe audio.mp3 --word-timestamps --format srt` | wav2vec2 aligned (~10ms) |
| **Speaker diarization** | `./scripts/transcribe audio.mp3 --diarize` | Requires pyannote.audio |
| **Translate → English** | `./scripts/transcribe audio.mp3 --translate` | Any language → English |
| **Stream output** | `./scripts/transcribe audio.mp3 --stream` | Live segments as transcribed |
| **Clip time range** | `./scripts/transcribe audio.mp3 --clip-timestamps "30,60"` | Only 30s–60s |
| **Denoise + normalize** | `./scripts/transcribe audio.mp3 --denoise --normalize` | Clean up noisy audio first |
| **Reduce hallucination** | `./scripts/transcribe audio.mp3 --hallucination-silence-threshold 1.0` | Skip hallucinated silence |
| **YouTube/URL** | `./scripts/transcribe https://youtube.com/watch?v=...` | Auto-downloads via yt-dlp |
| **Batch process** | `./scripts/transcribe *.mp3 -o ./transcripts/` | Output to directory |
| **Batch with skip** | `./scripts/transcribe *.mp3 --skip-existing -o ./out/` | Resume interrupted batches |
| **Domain terms** | `./scripts/transcribe audio.mp3 --initial-prompt 'Kubernetes gRPC'` | Boost rare terminology |
| **Hotwords boost** | `./scripts/transcribe audio.mp3 --hotwords 'JIRA Kubernetes'` | Bias decoder toward specific words |
| **Prefix conditioning** | `./scripts/transcribe audio.mp3 --prefix 'Good morning,'` | Seed the first segment with known opening words |
| **Pin model version** | `./scripts/transcribe audio.mp3 --revision v1.2.0` | Reproducible transcription with a pinned revision |
| **Debug library logs** | `./scripts/transcribe audio.mp3 --log-level debug` | Show faster_whisper internal logs |
| **Turbo model** | `./scripts/transcribe audio.mp3 -m turbo` | Alias for large-v3-turbo |
| **Faster English** | `./scripts/transcribe audio.mp3 --model distil-medium.en -l en` | English-only, 6.8x faster |
| **Maximum accuracy** | `./scripts/transcribe audio.mp3 --model large-v3 --beam-size 10` | Full model |
| **JSON output** | `./scripts/transcribe audio.mp3 --format json -o out.json` | Programmatic access with stats |
| **Filter noise** | `./scripts/transcribe audio.mp3 --min-confidence 0.6` | Drop low-confidence segments |
| **Hybrid quantization** | `./scripts/transcribe audio.mp3 --compute-type int8_float16` | Save VRAM, minimal quality loss |
| **Reduce batch size** | `./scripts/transcribe audio.mp3 --batch-size 4` | If OOM on GPU |
| **TSV output** | `./scripts/transcribe audio.mp3 --format tsv -o out.tsv` | OpenAI Whisper–compatible TSV |
| **Fix hallucinations** | `./scripts/transcribe audio.mp3 --temperature 0.0 --no-speech-threshold 0.8` | Lock temperature + skip silence |
| **Tune VAD sensitivity** | `./scripts/transcribe audio.mp3 --vad-threshold 0.6 --min-silence-duration 500` | Tighter speech detection |
| **Known speaker count** | `./scripts/transcribe meeting.wav --diarize --min-speakers 2 --max-speakers 3` | Constrain diarization |
| **Subtitle word wrapping** | `./scripts/transcribe audio.mp3 --format srt --word-timestamps --max-words-per-line 8` | Split long cues |
| **Private/gated model** | `./scripts/transcribe audio.mp3 --hf-token hf_xxx` | Pass token directly |
| **Show version** | `./scripts/transcribe --version` | Print faster-whisper version |
| **Upgrade in-place** | `./setup.sh --update` | Upgrade without full reinstall |
| **System check** | `./setup.sh --check` | Verify GPU, Python, ffmpeg, venv, yt-dlp, pyannote |
| **Detect language only** | `./scripts/transcribe audio.mp3 --detect-language-only` | Fast language ID, no transcription |
| **Detect language JSON** | `./scripts/transcribe audio.mp3 --detect-language-only --format json` | Machine-readable language detection |
| **LRC subtitles** | `./scripts/transcribe audio.mp3 --format lrc -o lyrics.lrc` | Timed lyrics format for music players |
| **ASS subtitles** | `./scripts/transcribe audio.mp3 --format ass -o subtitles.ass` | Advanced SubStation Alpha (Aegisub, mpv, VLC) |
| **Merge sentences** | `./scripts/transcribe audio.mp3 --format srt --merge-sentences` | Join fragments into sentence chunks |
| **Stats sidecar** | `./scripts/transcribe audio.mp3 --stats-file stats.json` | Write perf stats JSON after transcription |
| **Batch stats** | `./scripts/transcribe *.mp3 --stats-file ./stats/` | One stats file per input in dir |
| **Template naming** | `./scripts/transcribe audio.mp3 -o ./out/ --output-template "{stem}_{lang}.{ext}"` | Custom batch output filenames |
| **Stdin input** | `ffmpeg -i input.mp4 -f wav - \| ./scripts/transcribe -` | Pipe audio directly from stdin |
| **Custom model dir** | `./scripts/transcribe audio.mp3 --model-dir ~/my-models` | Custom HuggingFace cache dir |
| **Local model** | `./scripts/transcribe audio.mp3 -m ./my-model-ct2` | CTranslate2 model dir |
| **HTML transcript** | `./scripts/transcribe audio.mp3 --format html -o out.html` | Confidence-colored |
| **Burn subtitles** | `./scripts/transcribe video.mp4 --burn-in output.mp4` | Requires ffmpeg + video input |
| **Name speakers** | `./scripts/transcribe audio.mp3 --diarize --speaker-names "Alice,Bob"` | Replaces SPEAKER_1/2 |
| **Filter hallucinations** | `./scripts/transcribe audio.mp3 --filter-hallucinations` | Removes artifacts |
| **Keep temp files** | `./scripts/transcribe https://... --keep-temp` | For URL re-processing |
| **Parallel batch** | `./scripts/transcribe *.mp3 --parallel 4 -o ./out/` | CPU multi-file |
| **RTX 3070 recommended** | `./scripts/transcribe audio.mp3 --compute-type int8_float16` | Saves ~1GB VRAM, minimal quality loss |
| **CPU thread count** | `./scripts/transcribe audio.mp3 --threads 8` | Force CPU thread count (default: auto) |
| **Podcast RSS (latest 5)** | `./scripts/transcribe --rss https://feeds.example.com/podcast.xml` | Downloads & transcribes newest 5 episodes |
| **Podcast RSS (all episodes)** | `./scripts/transcribe --rss https://... --rss-latest 0 -o ./episodes/` | All episodes, one file each |
| **Podcast + SRT subtitles** | `./scripts/transcribe --rss https://... --format srt -o ./subs/` | Subtitle all episodes |
| **Retry on failure** | `./scripts/transcribe *.mp3 --retries 3 -o ./out/` | Retry up to 3× with backoff on error |
| **CSV output** | `./scripts/transcribe audio.mp3 --format csv -o out.csv` | Spreadsheet-ready with header row; properly quoted |
| **CSV with speakers** | `./scripts/transcribe audio.mp3 --diarize --format csv -o out.csv` | Adds speaker column |
| **Language map (inline)** | `./scripts/transcribe *.mp3 --language-map "interview*.mp3=en,lecture.wav=fr"` | Per-file language in batch |
| **Language map (JSON)** | `./scripts/transcribe *.mp3 --language-map @langs.json` | JSON file: {"pattern": "lang"} |
| **Batch with ETA** | `./scripts/transcribe *.mp3 -o ./out/` | Automatic ETA shown for each file in batch |
| **TTML subtitles** | `./scripts/transcribe audio.mp3 --format ttml -o subtitles.ttml` | Broadcast-standard DFXP/TTML (Netflix, BBC, Amazon) |
| **TTML with speaker labels** | `./scripts/transcribe audio.mp3 --diarize --format ttml -o subtitles.ttml` | Speaker-labeled TTML |
| **Search transcript** | `./scripts/transcribe audio.mp3 --search "keyword"` | Find timestamps where keyword appears |
| **Search to file** | `./scripts/transcribe audio.mp3 --search "keyword" -o results.txt` | Save search results |
| **Fuzzy search** | `./scripts/transcribe audio.mp3 --search "aproximate" --search-fuzzy` | Approximate/partial matching |
| **Detect chapters** | `./scripts/transcribe audio.mp3 --detect-chapters` | Auto-detect chapters from silence gaps |
| **Chapter gap tuning** | `./scripts/transcribe audio.mp3 --detect-chapters --chapter-gap 5` | Chapters on gaps ≥5s (default: 8s) |
| **Chapters to file** | `./scripts/transcribe audio.mp3 --detect-chapters --chapters-file ch.txt` | Save YouTube-format chapter list |
| **Chapters JSON** | `./scripts/transcribe audio.mp3 --detect-chapters --chapter-format json` | Machine-readable chapter list |
| **Export speaker audio** | `./scripts/transcribe audio.mp3 --diarize --export-speakers ./speakers/` | Save each speaker's audio to separate WAV files |
| **Multi-format output** | `./scripts/transcribe audio.mp3 --format srt,text -o ./out/` | Write SRT + TXT in one pass |
| **Remove filler words** | `./scripts/transcribe audio.mp3 --clean-filler` | Strip um/uh/er/ah/hmm and discourse markers |
| **Left channel only** | `./scripts/transcribe audio.mp3 --channel left` | Extract left stereo channel before transcribing |
| **Right channel only** | `./scripts/transcribe audio.mp3 --channel right` | Extract right stereo channel |
| **Max chars per line** | `./scripts/transcribe audio.mp3 --format srt --max-chars-per-line 42` | Character-based subtitle wrapping |
| **Detect paragraphs** | `./scripts/transcribe audio.mp3 --detect-paragraphs` | Insert paragraph breaks in text output |
| **Paragraph gap tuning** | `./scripts/transcribe audio.mp3 --detect-paragraphs --paragraph-gap 5.0` | Tune gap threshold (default 3.0s) |
## Model Selection
Choose the right model for your needs:
```dot
digraph model_selection {
rankdir=LR;
node [shape=box, style=rounded];
start [label="Start", shape=doublecircle];
need_accuracy [label="Need maximum\naccuracy?", shape=diamond];
multilingual [label="Multilingual\ncontent?", shape=diamond];
resource_constrained [label="Resource\nconstraints?", shape=diamond];
large_v3 [label="large-v3\nor\nlarge-v3-turbo", style="rounded,filled", fillcolor=lightblue];
large_turbo [label="large-v3-turbo", style="rounded,filled", fillcolor=lightblue];
distil_large [label="distil-large-v3.5\n(default)", style="rounded,filled", fillcolor=lightgreen];
distil_medium [label="distil-medium.en", style="rounded,filled", fillcolor=lightyellow];
distil_small [label="distil-small.en", style="rounded,filled", fillcolor=lightyellow];
start -> need_accuracy;
need_accuracy -> large_v3 [label="yes"];
need_accuracy -> multilingual [label="no"];
multilingual -> large_turbo [label="yes"];
multilingual -> resource_constrained [label="no (English)"];
resource_constrained -> distil_small [label="mobile/edge"];
resource_constrained -> distil_medium [label="some limits"];
resource_constrained -> distil_large [label="no"];
}
```
### Model Table
#### Standard Models (Full Whisper)
| Model | Size | Speed | Accuracy | Use Case |
| ---------------------- | ----- | --------- | --------- | ---------------------------------- |
| `tiny` / `tiny.en` | 39M | Fastest | Basic | Quick drafts |
| `base` / `base.en` | 74M | Very fast | Good | General use |
| `small` / `small.en` | 244M | Fast | Better | Most tasks |
| `medium` / `medium.en` | 769M | Moderate | High | Quality transcription |
| `large-v1/v2/v3` | 1.5GB | Slower | Best | Maximum accuracy |
| `large-v3-turbo` | 809M | Fast | Excellent | High accuracy (slower than distil) |
#### Distilled Models (~6x Faster, ~1% WER difference)
| Model | Size | Speed vs Standard | Accuracy | Use Case |
| ----------------------- | ---- | ----------------- | --------- | ---------------------------------- |
| **`distil-large-v3.5`** | 756M | ~6.3x faster | 7.08% WER | **Default, best balance** |
| `distil-large-v3` | 756M | ~6.3x faster | 7.53% WER | Previous default |
| `distil-large-v2` | 756M | ~5.8x faster | 10.1% WER | Fallback |
| `distil-medium.en` | 394M | ~6.8x faster | 11.1% WER | English-only, resource-constrained |
| `distil-small.en` | 166M | ~5.6x faster | 12.1% WER | Mobile/edge devices |
`.en` models are English-only and slightly faster/better for English content.
> **Note for distil models:** HuggingFace recommends disabling `condition_on_previous_text` for all distil models to prevent repetition loops. The script **auto-applies** `--no-condition-on-previous-text` whenever a `distil-*` model is detected. Pass `--condition-on-previous-text` to override if needed.
## Custom & Fine-tuned Models
WhisperModel accepts local CTranslate2 model directories and HuggingFace repo names — no code changes needed.
### Load a local CTranslate2 model
```bash
./scripts/transcribe audio.mp3 --model /path/to/my-model-ct2
```
### Convert a HuggingFace model to CTranslate2
```bash
pip install ctranslate2
ct2-transformers-converter \
--model openai/whisper-large-v3 \
--output_dir whisper-large-v3-ct2 \
--copy_files tokenizer.json preprocessor_config.json \
--quantization float16
./scripts/transcribe audio.mp3 --model ./whisper-large-v3-ct2
```
### Load a model by HuggingFace repo name (auto-downloads)
```bash
./scripts/transcribe audio.mp3 --model username/whisper-large-v3-ct2
```
### Custom model cache directory
By default, models are cached in `~/.cache/huggingface/`. Use `--model-dir` to override:
```bash
./scripts/transcribe audio.mp3 --model-dir ~/my-models
```
## Setup
### Linux / macOS / WSL2
```bash
# Base install (creates venv, installs deps, auto-detects GPU)
./setup.sh
# With speaker diarization support
./setup.sh --diarize
```
Requirements:
- Python 3.10+
- ffmpeg is **not required** for basic transcription — PyAV (bundled with faster-whisper) handles audio decoding. ffmpeg is only needed for `--burn-in`, `--normalize`, and `--denoise`.
- Optional: yt-dlp (for URL/YouTube input)
- Optional: pyannote.audio (for `--diarize`, installed via `setup.sh --diarize`)
### Platform Support
| Platform | Acceleration | Speed |
| ---------------------- | ------------ | ---------------- |
| **Linux + NVIDIA GPU** | CUDA | ~20x realtime 🚀 |
| **WSL2 + NVIDIA GPU** | CUDA | ~20x realtime 🚀 |
| macOS Apple Silicon | CPU\* | ~3-5x realtime |
| macOS Intel | CPU | ~1-2x realtime |
| Linux (no GPU) | CPU | ~1x realtime |
\*faster-whisper uses CTranslate2 which is CPU-only on macOS, but Apple Silicon is fast enough for practical use.
### GPU Support (IMPORTANT!)
The setup script auto-detects your GPU and installs PyTorch with CUDA. **Always use GPU if available** — CPU transcription is extremely slow.
| Hardware | Speed | 9-min video |
| -------------- | -------------- | ----------- |
| RTX 3070 (GPU) | ~20x realtime | ~27 sec |
| CPU (int8) | ~0.3x realtime | ~30 min |
> **RTX 3070 tip**: Use `--compute-type int8_float16` for hybrid quantization — saves ~1GB VRAM with minimal quality loss. Ideal for running diarization alongside transcription.
If setup didn't detect your GPU, manually install PyTorch with CUDA:
```bash
# For CUDA 12.x
uv pip install --python .venv/bin/python torch --index-url https://download.pytorch.org/whl/cu121
# For CUDA 11.x
uv pip install --python .venv/bin/python torch --index-url https://download.pytorch.org/whl/cu118
```
- **WSL2 users**: Ensure you have the [NVIDIA CUDA drivers for WSL](https://docs.nvidia.com/cuda/wsl-user-guide/) installed on Windows
## Usage
```bash
# Basic transcription
./scripts/transcribe audio.mp3
# SRT subtitles
./scripts/transcribe audio.mp3 --format srt -o subtitles.srt
# WebVTT subtitles
./scripts/transcribe audio.mp3 --format vtt -o subtitles.vtt
# Transcribe from YouTube URL
./scripts/transcribe https://youtube.com/watch?v=dQw4w9WgXcQ --language en
# Speaker diarization
./scripts/transcribe meeting.wav --diarize
# Diarized VTT subtitles
./scripts/transcribe meeting.wav --diarize --format vtt -o meeting.vtt
# Prime with domain terminology
./scripts/transcribe lecture.mp3 --initial-prompt "Kubernetes, gRPC, PostgreSQL, NGINX"
# Batch process a directory
./scripts/transcribe ./recordings/ -o ./transcripts/
# Batch with glob, skip already-done files
./scripts/transcribe *.mp3 --skip-existing -o ./transcripts/
# Filter low-confidence segments
./scripts/transcribe noisy-audio.mp3 --min-confidence 0.6
# JSON output with full metadata
./scripts/transcribe audio.mp3 --format json -o result.json
# Specify language (faster than auto-detect)
./scripts/transcribe audio.mp3 --language en
```
## Options
```
Input:
AUDIO Audio file(s), directory, glob pattern, or URL
Accepts: mp3, wav, m4a, flac, ogg, webm, mp4, mkv, avi, wma, aac
URLs auto-download via yt-dlp (YouTube, direct links, etc.)
Model & Language:
-m, --model NAME Whisper model (default: distil-large-v3.5; "turbo" = large-v3-turbo)
--revision REV Model revision (git branch/tag/commit) to pin a specific version
-l, --language CODE Language code, e.g. en, es, fr (auto-detects if omitted)
--initial-prompt TEXT Prompt to condition the model (terminology, formatting style)
--prefix TEXT Prefix to condition the first segment (e.g. known starting words)
--hotwords WORDS Space-separated hotwords to boost recognition
--translate Translate any language to English (instead of transcribing)
--multilingual Enable multilingual/code-switching mode (helps smaller models)
--hf-token TOKEN HuggingFace token for private/gated models and diarization
--model-dir PATH Custom model cache directory (default: ~/.cache/huggingface/)
Output Format:
-f, --format FMT text | json | srt | vtt | tsv | lrc | html | ass | ttml (default: text)
Accepts comma-separated list: --format srt,text writes both in one pass
Multi-format requires -o <dir> when saving to files
--word-timestamps Include word-level timestamps (wav2vec2 aligned automatically)
--stream Output segments as they are transcribed (disables diarize/alignment)
--max-words-per-line N For SRT/VTT, split segments into sub-cues of at most N words
--max-chars-per-line N For SRT/VTT/ASS/TTML, split lines so each fits within N characters
Takes priority over --max-words-per-line when both are set
--clean-filler Remove hesitation fillers (um, uh, er, ah, hmm, hm) and discourse markers
(you know, I mean, you see) from transcript text. Off by default.
--detect-paragraphs Insert paragraph breaks (blank lines) in text output at natural boundaries.
A new paragraph starts when: silence gap ≥ --paragraph-gap, OR the previous
segment ends a sentence AND the gap ≥ 1.5s.
--paragraph-gap SEC Minimum silence gap in seconds to start a new paragraph (default: 3.0).
Used with --detect-paragraphs.
--channel {left,right,mix}
Stereo channel to transcribe: left (c0), right (c1), or mix (default: mix).
Extracts the channel via ffmpeg before transcription. Requires ffmpeg.
--merge-sentences Merge consecutive segments into sentence-level chunks
(improves SRT/VTT readability; groups by terminal punctuation or >2s gap)
-o, --output PATH Output file or directory (directory for batch mode)
--output-template TEMPLATE
Batch output filename template. Variables: {stem}, {lang}, {ext}, {model}
Example: "{stem}_{lang}.{ext}" → "interview_en.srt"
Inference Tuning:
--beam-size N Beam search size; higher = more accurate but slower (default: 5)
--temperature T Sampling temperature or comma-separated fallback list, e.g.
'0.0' or '0.0,0.2,0.4' (default: faster-whisper's schedule)
--no-speech-threshold PROB
Probability threshold to mark segments as silence (default: 0.6)
--batch-size N Batched inference batch size (default: 8; reduce if OOM)
--no-vad Disable voice activity detection (on by default)
--vad-threshold T VAD speech probability threshold (default: 0.5)
--vad-neg-threshold T VAD negative threshold for ending speech (default: auto)
--vad-onset T Alias for --vad-threshold (legacy)
--vad-offset T Alias for --vad-neg-threshold (legacy)
--min-speech-duration MS Minimum speech segment duration in ms (default: 0)
--max-speech-duration SEC Maximum speech segment duration in seconds (default: unlimited)
--min-silence-duration MS Minimum silence before splitting a segment in ms (default: 2000)
--speech-pad MS Padding around speech segments in ms (default: 400)
--no-batch Disable batched inference (use standard WhisperModel)
--hallucination-silence-threshold SEC
Skip silent sections where model hallucinates (e.g. 1.0)
--no-condition-on-previous-text
Don't condition on previous text (reduces repetition/hallucination loops;
auto-enabled for distil models per HuggingFace recommendation)
--condition-on-previous-text
Force-enable conditioning on previous text (overrides auto-disable for distil models)
--compression-ratio-threshold RATIO
Filter segments above this compression ratio (default: 2.4)
--log-prob-threshold PROB
Filter segments below this avg log probability (default: -1.0)
--max-new-tokens N Maximum tokens per segment (prevents runaway generation)
--clip-timestamps RANGE
Transcribe specific time ranges: '30,60' or '0,30;60,90' (seconds)
--progress Show transcription progress bar
--best-of N Candidates when sampling with non-zero temperature (default: 5)
--patience F Beam search patience factor (default: 1.0)
--repetition-penalty F Penalty for repeated tokens (default: 1.0)
--no-repeat-ngram-size N Prevent n-gram repetitions of this size (default: 0 = off)
Advanced Inference:
--no-timestamps Output text without timing info (faster; incompatible with
--word-timestamps, --format srt/vtt/tsv, --diarize)
--chunk-length N Audio chunk length in seconds for batched inference (default: auto)
--language-detection-threshold T
Confidence threshold for language auto-de原文较长,这里只显示前 39 KB,完整内容请到来源页面查看。
通过同一个AI创作工具完成AI图片、视频、音乐与语音创作,查找公开社交数据,编辑视觉成果,并集中查看和管理生成资产。Beatra 全能创作支持AI图片生成(文生图)、AI视频生成(图生视频)和AI视频编辑,通过AI音乐生成把想法做成歌曲或纯音乐,也可用AI语音生成(文字转语音)、多语言旁白和AI语音克隆完成跨媒介内容;还能查找抖音数据、小红书数据,以及 YouTube、TikTok、Instagram 与 X 上的公开帖子、笔记、评论、账号和趋势,把社交媒体数据用于竞品调研和创作。它是一款面向AI内容创作与多模态内容创作的AI创作套件,用起来就是一个把图片、视频、音乐与语音收在同一处的AI创作助手,适用于社交内容、产品展示、营销素材、课程、播客、短片和跨媒介项目;还能在不同媒介间复用原始素材、已有成果和查到的公开资料,查看制作进度、返回文件、实际用量和积分记录。
用一张人像和一段短脚本或语音,制作有明确讲述内容的数字人口播视频。可使用现成录音,也可以先选择音色,把脚本生成旁白,让画面中的人物随声音完成稳定、自然的口播。适合 AI 主播、虚拟主播视频、数字人讲解、让照片说话、照片说话视频、产品讲解视频、数字人培训视频、公告和社交口播,并在交付后检查人物一致性、声音清晰度、口型同步和动作稳定性。一节课的讲解也能做成课程口播。
ContextWeave 架构图一键生成
将一段活动描述、一张产品照片或品牌参考图,转化为抓人眼球的活动海报、宣传 banner 或社交媒体配图。这款 AI海报设计与制作工具生成图片丰富的海报,在海报排版上搭配清晰视觉层级、按品类匹配的风格和预留文字区域,覆盖活动、音乐节、电影、产品发布和促销场景。你可以把它当作 AI宣传单制作工具来制作促销宣传单,当作活动海报设计工具,当作电影海报生成器,当作音乐节海报设计工具,当作产品发布会海报的 banner设计 AI,或当作平面设计海报工作室。从活动主题出发,用品牌风格参考图变换产品或场景照片,或在已认可草稿上做定向精修,完成可印刷或可分享的成品。
抖音短视频双路径生产体系(单 Skill 版)。Path A[付费·高质量]:策略大脑→(可选:素材混剪/AI高光提取)→脚本优化→AI生图/素材准备→配音渲染(百炼/GPT-SoVITS/CosyVoice/VoxCPM2+剪映/Remotion/ffmpeg/数字人)。Path B[免费·动画风]:采集热点→脚本→edge-tts免费配音+HyperFrames开源渲染→MP4,零云费开箱快。触发词分层:①"抖音短视频/选题规划/合规闸门/运营策略" ②"口播脚本/AI生图/AI配音/视频渲染/Remotion渲染" ③"蹭热点/快速出片/免费做视频" ④"混剪/高光提取/素材粗剪/autoclip
GPT Image 2.5生图是一款 AI 工具,用 GPT-Image-2.5 和 ChatGPT Images 2.5 出一张图。GPT Image 2.5 Studio 支持 gpt-image-2.5-flare 与 gpt-image-2.5-sunburst,可文生图或参考图生图,做中文海报、商品图、信息图和社媒画面。
用 MiniMax H3 这款 AI视频生成器创作高质感 2K AI 视频。从文字创意出发完成文生视频,让静态图片自然动起来,用首尾画面做首尾帧视频的完整转场,也可以结合图片、视频和音频参考,以多模态视频生成塑造统一场景。适合 AI 广告视频、电商产品视频、品牌宣传视频、动态海报、游戏 UI、电影片头和社交短视频创作,形成电影感画面与同步的原生立体声。使用 Beatra 账号即可开始,创作进度与成片集中管理。
输入品牌名称、行业方向或参考图,生成专业 AI Logo、品牌标识和应用图标。这款 AI logo 设计工具从你的品牌简报出发——公司名、行业气质、配色偏好——先探索多个商标设计方向,再用精确品牌色、简洁几何和可缩放构图精修最强概念。制作企业 logo、创业公司标志、字母标、徽章、组合标识和 VI设计中的主标,从网站 favicon 到门头招牌都清晰可辨。从文字描述出发,将已有草图和品牌素材转化为精致标识,或在已认可草稿上做定向精修。
数据月份:2026-10 · 下载/星标数聚合自公开技能注册库(ClawHub、SkillHub),按月更新。口径说明见数据方法。