// the find
wxbool/video-srt-windows
这是一个可以识别视频语音自动生成字幕SRT文件的开源 Windows-GUI 软件工具。
VideoSrt is a Windows GUI tool that extracts audio from video/audio files and generates SRT subtitle files using Alibaba Cloud's speech recognition API. It targets Chinese-speaking users who need to batch-generate subtitles for Mandarin or English content. The repo is abandoned — last commit was March 2023 and the author has since launched a paid Pro version.
Practical pipeline: uses ffmpeg to strip audio, uploads to Alibaba Cloud OSS, calls their file transcription API, then writes SRT/LRC/plain text — the architecture is straightforward and auditable. Supports multiple translation engines (Baidu, Tencent) and bilingual subtitle output. Batch processing is a first-class feature, not an afterthought. Go + lxn/walk means a single self-contained executable with no runtime dependencies beyond ffmpeg.
Hard dependency on Alibaba Cloud OSS and their NLS API means you're paying per minute and handing your audio to a third party — there's no local/offline option. The author's own README promotes a paid successor (VideoSrt Pro) and a paid SaaS, which signals this codebase is a dead end. Windows-only by design (lxn/walk), so non-Windows users are out of luck. No tests anywhere in the tree, and the upgrade path is manual file-copying — there's no installer or auto-update mechanism.