The "tts" topic on GitHub groups 36 open-source projects in the DeepSeek Harness (DSH) ecosystem, led by dsh-voice-ai-girlfriend with 73 GitHub stars. dsh-voice-ai-girlfriend — Voice AI girlfriend (Voice AI girlfriend for DeepSeek Harness): Whisper voice input + Qwen3-TTS voice cloning + sentence-level streaming read-aloud + digital human animation window. Every project here is indexed by DSH Universe with live GitHub data — stars, activity and install status — so you can compare and install directly.
Voice-first session loop for DeepSeek Harness: a composer microphone button with browser/local speech-to-text (Web Speech, FunASR, whisper.cpp), a speak tool for text-to-speech replies (browser, edge-tts, piper), event announcements with mute, and speak-to-interrupt.
One tool = all MiniMax multimodal capabilities: DSH text-only models see images/draw images/generate video/speak/sing/cover/search/check quota | One mmx_bridge tool = all MiniMax multimodal (VLM/image/video/speech/music/cover/search/quota) for DeepSeek Harness (DSH)
DeepSeek Harness plugin: notification outlet — the agent reaches the user via desktop notifications / Chinese voice announcements / alert sounds (long task finished, error, calling the user back). Zero dependencies on Windows.
Voice notes in, spoken answers out — dictate audio that becomes user messages (transcribe), have the agent read replies aloud (speak), and leave walk-away narration on long headless runs. Local-first: plain audio files under ~/.dsh/voice/.
Fish Audio TTS plugin for DeepSeek Harness — per-reply read-aloud, auto-read toggle, BYOK, third-party. TTS plugin: per-reply read-aloud, auto-read; only supports Fish Audio API, bring your own Key; third-party unofficial.
Give a DeepSeek Harness agent a voice it owns — offer_call rings the human (answer/decline/later); local TTS via CrispASR + Qwen3-TTS CustomVoice, 9 speakers, 2 Chinese dialects. Local-first, offline-capable.
DSH full-duplex voice conversation mode: streaming zipformer2 recognition into an editable draft, optional wake word, Edge TTS sentence-by-sentence reading + real-time subtitles, barge-in on speech, no API Key needed. Full-duplex voice mode for DeepSeek Harness, no API key.
Captain Call - WeChat-style voice calls with your AgentTeams agents in DeepSeek Harness: GINKA desktop assistant, Kokoro open-source Chinese TTS, contact book with per-member voice picker. | Captain Call: WeChat-style voice call plugin with your Agent teammates
Speech plugin for DeepSeek Harness: per-message voice playback, auto-announce, and dictation via mic — cloud TTS/ASR on Alibaba DashScope or Volcengine (freely combinable), with Web Speech API fallback.
Chat with, monitor, and approve your DSH (DeepSeek Harness) agents from WeChat over the clawbot iLink gateway: two-way text/images/voice/files/video, native vision or OCR, context-rotation policies, reminders, and a standalone admin console.
Let an AI assistant drive STUDIO NEUTRINO (free Japanese AI singing synthesis engine) to sing: give lyrics and melody, and it auto-generates a WAV of the character singing. Bring your own free NEUTRINO engine and voice banks; the plugin includes official download guidance.
Voice reading plugin for DeepSeek Harness: streaming TTS reading of AI replies, plus fun voice-phrase feedback during thinking/waiting. Edge TTS works out of the box, supports any OpenAI-compatible cloud engine.
Spoken status alerts for DeepSeek Harness where clip LENGTH encodes urgency: a 2s approval chime vs an 8s failure announcement. Eight scenes, zero third-party dependencies on Windows 10/11. · DSH 语音状态提示:用时长编码紧急度,零第三方依赖。
Really Love You: local smart video butler Skill —— say 「Really Love You」 and it takes over the entire editing pipeline from footage analysis to final delivery