VideoCaptioner

Why we included this project

VideoCaptioner is a practical end-to-end tool for anyone who regularly produces subtitled video: course creators, translators, YouTubers, or teams localizing internal training footage. It takes a video file through the whole pipeline, from speech-to-text transcription and sentence segmentation to LLM-based cleanup and translation, then burns the result into the video as soft or hard subtitles. The free tiers, Bijian speech recognition and Bing/Google translation, work out of the box with no API key, so you can test the workflow before deciding whether to add an LLM provider for better optimization and translation quality. The tool ships as both a desktop GUI and a CLI, and the CLI supports batch processing plus a single command that runs the full pipeline, which makes it easy to slot into existing production scripts. Word-level timestamps and VAD detection help keep timing accurate, and low token consumption keeps LLM costs modest for large batches.