video-to-text
Skill0 starsChanged 49 days agoMITView on GitHub
提取视频语音并转成中文文字稿。当用户给出抖音、B站(bilibili/b23.tv)、YouTube、快手、视频号等常见视频网站的链接,想要"提取视频里说话的内容""视频转文字""语音转文字""视频字幕提取"时,必须使用本技能。即使用户只发了一个视频链接没说要做什么,也应主动用本技能识别并转写其中的说话内容。
- Installs packages
5 other projects use a version of this.
Adopted by
Most recently taken up by
- lgldlk/lgldlk-agent-skillstheir own version, first seen Oct 13, 2026
- jiawood2006/video-to-texttheir own version, first seen Oct 13, 2026
- jiawood2006/hermes-skillstheir own version, first seen Aug 24, 2026
- Lotus-7/video-to-text-skilltheir own version, first seen May 8, 2026
- yfge/video-skills-suitetheir own version, first seen Feb 28, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- lgldlk/lgldlk-agent-skills★ 3same name
- jiawood2006/hermes-skills★ 2same name
- yfge/video-skills-suite★ 1same name
- jiawood2006/video-to-text★ 0same name
- Lotus-7/video-to-text-skill★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
Files
视频语音转文字
把常见视频网站的链接一键转成中文文字稿(带时间轴 + 纯文字两版)。
支持的站点与下载方式
| 站点 | 链接形式 | 下载方式 |
| 抖音 | v.douyin.com 短链 / douyin.com/video/ID | yt-dlp 会失败(需要新鲜 Cookie),自动回退无头 Chromium 捕获 |
| B站 | b23.tv 短链 / bilibili.com/video/BV号 | yt-dlp 直接下载 |
| YouTube | youtu.be / youtube.com/watch | yt-dlp 直接下载 |
| 快手、其他 | 短链或长链 | 先 yt-dlp,失败后 Chromium 回退 |
用法
python <skill目录>/scripts/video_to_text.py "<视频链接>" [输出目录] [--model small] [--keep-media]- 输出目录缺省为当前目录,产出
文字稿.txt和meta.json(含标题、站点信息) --model:whisper 模型大小,base更快、small(默认)准确率更高、medium更准但更慢--keep-media:保留下载的音视频中间文件(默认转写后清理)- 直接管道输出文字稿,同时在回复中完整引用给用户
内部流程(脚本已自动完成,仅在排错时需要了解)
- 展开短链接(跟随 302)
- 站点识别与 URL 规范化(提取视频 ID,去掉跟踪参数)
yt-dlp -x提取音频;失败且为抖音类站点时,用无头 Chromium 打开页面捕获媒体直链(优先纯音频轨,体积小得多)- 用 imageio-ffmpeg 的完整版 ffmpeg(不是 Playwright 自带的精简版——那个没有 mp4 解码器,会报 "Invalid data found when processing input")统一转成 16kHz 单声道 WAV
- faster-whisper 中文转写(CPU int8,VAD 过滤静音)
依赖检查
脚本依赖:requests、imageio-ffmpeg、faster-whisper、yt-dlp、playwright。 首次使用前确认:
python -c "import requests, imageio_ffmpeg, faster_whisper, yt_dlp, playwright"缺什么装什么(pip install <包名>)。Chromium 在 %LOCALAPPDATA%\ms-playwright\ 下自动发现;
没有的话 python -m playwright install chromium。
长视频处理
单次转写受执行环境超时限制(约 5 分钟)。音频超过约 8 分钟时,改用分段策略:
- 下载音频后转 16kHz WAV
- 用 ffmpeg 按 6~7 分钟切段(段间留 10 秒重叠,如
-ss 0/-t 400、-ss 390/-t 400……) - 逐段调用 faster-whisper,时间戳加上段的偏移量
- 合并时丢弃每段开头约 12 秒(与上一段重叠的部分),再统一做术语校正
英文术语(DeFi、NFT、GameFi、DApps 等)在中文语音里极易被听错,后处理时务必按上下文校正。
结果后处理原则
- 国学、方言、专业术语多的视频,small 模型会听错同音字(如"运有变数"听成"孕有变数"、"言传身教"听成"言传深教")。结合视频标题和上下文语义校正明显的同音错字,并在交付时向用户说明做了哪些校正
- 长视频(>10 分钟)转写耗时会明显变长,可先用
--model base快速出稿 - 无语音或纯背景音乐的视频,转写结果为空,如实告知用户
- 始终用中文回复用户
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

