The latest articles on DEV Community by imper (@imper_7cde72b79d2529291ec). https://dev.to/imper_7cde72b79d2529291ec https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3812341%2F4a3e352b-d42f-4efe-bb2c-23e8703e87bb.jpg https://dev.to/imper_7cde72b79d2529291ec en imper Sun, 08 Mar 2026 02:37:46 +0000 https://dev.to/imper_7cde72b79d2529291ec/building-an-offline-subtitle-extractor-with-whispercpp-and-electron-44k2 https://dev.to/imper_7cde72b79d2529291ec/building-an-offline-subtitle-extractor-with-whispercpp-and-electron-44k2 <p>I watch a lot of foreign language content - anime, K-dramas, tech talks - and getting subtitles was always a pain. Upload to a random website, hit the daily limit, try another one, or install Python and figure out whisper's CLI.</p> <p>So over the past few months I've been building a desktop app that handles the whole pipeline locally.</p> <h2> What it does </h2> <p>You drop a video file in, pick a whisper model size, and it spits out an SRT subtitle file. Optionally you can translate the subtitles using one of several engines.</p> <p>The speech-to-text runs via <strong>whisper.cpp</strong> so everything stays on your machine. No uploads, no API calls for the transcription part. If you have an NVIDIA GPU it automatically uses CUDA, otherwise it falls back to CPU - this was one of the trickier parts to get right.</p> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F0bixenk40upft6zx7vbk.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2F0bixenk40upft6zx7vbk.png" alt="App Screenshot" width="800" height="481"></a></p> <h2> Tech decisions </h2> <p>I went with <strong>Electron + Node.js</strong> because:</p> <ul> <li>Cross-platform (though I'm mainly targeting Windows right now)</li> <li>Easy to bundle whisper.cpp binaries and ffmpeg</li> <li>The UI is just HTML/CSS/JS so iteration is fast</li> </ul> <p>For whisper.cpp integration, the app spawns it as a child process with the right flags depending on whether CUDA is available. Model files (GGML format) auto-download on first run into a local <code>_models/</code> folder.</p> <h2> Translation engines </h2> <p>Translation is optional. Currently supported:</p> <ul> <li> <strong>MyMemory</strong> - free, no API key, ~50K chars/day</li> <li> <strong>DeepL</strong> - free tier 500K chars/month, needs API key</li> <li> <strong>OpenAI GPT</strong> - paid, good quality for nuanced translations</li> <li> <strong>Gemini</strong> - Google's API, generous free tier</li> </ul> <p>The app chunks subtitle text and sends it in batches to avoid rate limits. Each engine has its own quirks with language codes so there's a mapping layer.</p> <h2> v1.4.0 changes </h2> <p>Just pushed the latest update which adds:</p> <ul> <li>Automatic GPU/CPU fallback detection</li> <li>Bundled ffprobe-static (no more separate ffmpeg install)</li> <li>Better DeepL language code mapping</li> </ul> <h2> Try it out </h2> <p>It's packaged as a portable .exe - no install needed, just extract and run.</p> <p><strong>GitHub</strong>: <a href="https://github.com/Blue-B/WhisperSubTranslate" rel="noopener noreferrer">WhisperSubTranslate</a><br> <strong>License</strong>: GPL-3.0</p> <p>If you're working on something similar or have suggestions, I'd love to hear about it.</p> opensource javascript electron whisper