gemini-live-api-dev: Gemini Live API Real-Time Streaming Development Guide
gemini-live-api-dev is a development skill in Google Gemini's official gemini-skills repository. It is intended for building real-time, bidirectional audio/video/text streaming applications with the Gemini Live API over WebSocket, and supports migrating from 2.0/2.5/3.1 Live to Gemini 3.8. It covers VAD, extended thinking, asynchronous function calling, session management, ephemeral tokens, and connecting/streaming examples for google-genai and @google/genai (>=2.3.0). Install it with npx skills add. It is suitable for real-time voice assistants, multimodal Live sessions, and secure browser authentication scenarios.
Read Moredsh-voice-live: DeepSeek Harness Real-Time Full-Duplex Voice Plugin
`dsh-voice-live` is a plugin that provides real-time full-duplex voice capabilities for DeepSeek Harness, designed to simplify developers' integration of real-time voice conversations. It integrates Volcano Engine's streaming ASR and TTS, supporting features such as microphone capture, real-time subtitles, barge-in interaction, and automatic submission after silence. The plugin has not yet been officially published via npm and must be built and run from the monorepo in its GitHub repository. Due to missing dependencies and browser limitations, wake word functionality is currently limited to Safari, while Chrome is affected by an echo cancellation bug. Developers can refer to the source code to integrate and use it in the monorepo.
Read MoreDeepSeek Harness Real-Time Voice Plugin (dsh-realtime-voice)
The DeepSeek Harness plugin `dsh-realtime-voice` enables real-time voice interaction between browser audio and LLMs. The plugin supports dual routing for Alibaba Cloud Bailian and OpenAI, uses a same-origin WebSocket proxy to provide automatic speech recognition (ASR) and text-to-speech (TTS), and features streaming audio playback with a candidate interruption review mechanism. After installation, users can use the plugin by configuring the corresponding API key in Harness. Considerations include microphone permission limitations in the desktop Electron version, strict same-origin security policies, and a voiceprint feature intended only for suppressing background speech.
Read More