dsh-audio-dub
Run the following command in DeepSeek Harness:
dsh plugin install pinch-eng/dsh-audio-dub
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install pinch-eng/dsh-audio-dub in the DeepSeek Harness terminal or config to install this plugin; source is available at https://github.com/pinch-eng/dsh-audio-dub
About this plugin
An agent can translate your copy in a heartbeat, but the moment you ask it to hand back an English version of this Chinese video the toolchain falls apart. You either open a web app and upload by hand, or wire together ffmpeg, ASR, a translation model, TTS and an alignment pass yourself. dsh-audio-dub collapses all of that into a single tool call: hand the agent a local file path or a direct media URL, and it gets back a finished dub—transcription, translation, voice cloning and timing all handled server-side, with the original speaker's voice preserved through AI cloning. Large files stream through a presigned upload target, so even a two-gigabyte clip never lands fully in memory.
The plugin exposes three task-shaped tools. dub_media submits the job and can block until the result is ready; dub_status lets you poll a job that outlived its wait budget or refresh a 48-hour download link; dub_languages is a local, zero-cost lookup of the ten supported target languages (Chinese, English, Spanish, French, German, Portuguese, Russian, Japanese, Korean, Italian), with auto-detection available for the source language. Running out of wait time is not an error—the job keeps going server-side and the returned job_id can be picked up at any moment.
It is aimed at developers, content teams and multi-language product builds that want media localization inside an automated pipeline: no per-video web dashboard, no home-grown ffmpeg-plus-Whisper-plus-TTS glue. One sentence, one file, one finished dub. Priced at $0.50 per minute of media, with a 60-minute and 2 GB per-file cap.
Use Cases
- One-click conversion of Chinese tutorial videos into dubs in ten target languages
- Batch audio-video translation and timeline alignment inside automated pipelines
- Retaining the original speaker's voice for overseas content release
Best For
- Teams producing multi-language content at scale
- Developers building automated media pipelines
- Solo creators localizing videos and podcasts
Related Plugins
ModLens is a vision plugin for DeepSeek Harness that gives text-only models sight by reading images pasted directly into chat, with zero-config setup and multiple vision engines.
On-demand vision for text-only DeepSeek Harness agents: built-in free keyless vision chain and 14 vision tools, routing image turns as tool calls to vision models with pixel fidelity, no Python needed, one-command install.
Give text-only models in DeepSeek Harness eyes, enabling image Q&A, long-screenshot OCR, UI restoration, and GUI visual tasks.
An armor-breaking plugin for DeepSeek: rewrites instruction priority, expands refusal-blocked phrases, persists across sessions, and shows a green active indicator.