Real-time system audio translation for macOS — translate any audio (YouTube, podcasts, meetings) live on screen with OpenAI or Google Gemini. On-device speech recognition; optional low-latency realtime mode.

VoxGo：面向 PC 游戏玩家的开源实时语音翻译浮窗，支持系统声音捕获、Whisper 识别、游戏字幕和手机同步。

ACL 2026 Demo: Real-time bidirectional Korean-English speech translation over PSTN telephone calls. Server-side relay with dual-session echo gating. 555ms median latency, USD 0.28/min.

A Powerful Real-Time Chat Platform that breaks language barriers by automatically translating messages between users in different languages. Built using modern web technologies, this application enables seamless global communication with built-in voice accessibility for inclusive and user-friendly interaction.

Real‑time voice translation with offline/online ASR & multi‑backend translation.

Open-source Chrome extension for real-time, two-way AI voice translation in Google Meet, Zoom Web, and multilingual browser meetings.

Realtime speech transcription and translation plugin for macOS IINA

🎙 Realtime multilingual voice translation with live captions, group rooms, FFT waveform & subtitle-style UI — Qwen3-ASR + DeepSeek + OpenAI Realtime, FastAPI + Vanilla JS

Real-time multilingual voice translation rooms — speak and listen in your own language. Rust/Axum + Deepgram + Groq, Astro client. Deploys on Railway + Vercel.

OpenAI Realtime Translation PoC for YouTube live translation with WebRTC, Chrome tab audio, and Korean captions

Earshot: Fast real-time speech translation (60+ base languages and more with model switch) for group voice calls. With diarization via individual speaker pre-enrollment (15s) and long-running discussion context. Highly customizable & optimizable 100% local inference

Open-source Chrome MV3 extension for real-time meeting and video translation, built around BYOK provider keys and no audio/transcript storage.

Android 实时翻译工具：麦克风同传、应用内音频捕获与悬浮字幕

Real-time multilingual interpreter for live events on OpenAI gpt-realtime-translate — one speaker, attendees get translated audio + live captions, every event archived to Backblaze B2 (source audio, transcripts, VTT/SRT). FastAPI + Next.js B2 sample app.

A beginner-friendly Python prototype for real-time speech and text translation, using Hugging Face translation models, Google Speech Recognition, and text-to-speech demos.

Ultra-low latency real-time voice-to-voice translation agent built with Google Gemini Multimodal Live API (WebSocket)

Parley - realtime speech translation Android app (OpenAI gpt-realtime-translate, Kotlin/Compose, BYOK)