kizuna-ai-lab/sokuji — Live speech translation powered by on-device AI and cloud providers — OpenAI, Google Gemini, Palabra…

Description and use

kizuna-ai-lab/sokuji is an open‑source web application that provides real‑time speech translation by leveraging both on‑device AI models and a variety of cloud providers such as OpenAI, Google Gemini, Palabra.ai, Kizuna AI, and Volcengine. The platform captures audio, transcribes it, and instantly translates the text into multiple languages, offering a responsive frontend UI for seamless multilingual communication. It’s designed for developers who want a flexible, plug‑and‑play solution that can switch between local inference and external APIs depending on performance, cost, or privacy requirements.

Problem

kizuna-ai-lab/sokuji helps add AI capability without starting from a blank model stack.

Outcome

Medium: useful for prototypes or internal workflows, with dependency and maintenance checks before production.

Workflow

• prototype AI features
• build RAG or agent workflows
• evaluate model tooling

Evidence

Live speech translation powered by on-device AI and cloud providers — OpenAI, Google Gemini, Palabra.ai, Kizuna AI, Volcengine, and more

Source