dsh-plugin-tts

Reading through lengthy AI responses during extended interactions can often lead to visual fatigue. dsh-plugin-tts is designed to solve this exact pain point by bringing a seamless, dual-sided text-to-speech experience to DeepSeek Harness. Whether you need the free, out-of-the-box Microsoft Edge TTS for online synthesis or prefer highly personalized local RVC custom voices, this plugin empowers you to give your AI assistant the perfect voice.

Its standout feature is the gapless adaptive chunked streaming playback. Instead of waiting for an entire long response to be synthesized, the plugin converts and plays audio chunks progressively, using Web Audio for sample-accurate joins that completely eliminate awkward pauses and gaps. Combined with handy features like selected-text reading, a mini player, auto-read toggles, and Agent approval voice alerts, it significantly elevates the voice interaction experience.

This plugin is ideal for DeepSeek Harness users who want to rest their eyes and prefer listening to AI replies, as well as advanced users looking to clone their own voice or use specific character models via RVC. With zero-friction setup—thanks to one-click voice pack installations and a portable runtime—you can effortlessly unlock a smooth, immersive, and customized audio companion for your workflow.