HojoAI/Hojo-TTS-Light

Hojo-TTS-Light is a lightweight speech generation model series developed by HojoAI, currently comprising a 0.04B and a 0.08B model.

What it solves

Hojo-TTS-Light is designed to provide high-quality, bilingual (Chinese and English) text-to-speech synthesis that can run on low-power hardware. It addresses the high computational requirements typically associated with advanced TTS models, allowing for efficient deployment on CPUs, ordinary GPUs, and embedded edge devices.

How it works

The project uses a token-LM framework based on a speech tokenizer (consisting of an 18M encoder and a 30M decoder). It offers two model sizes: a 40M parameter model with 15 built-in voices, and an 80M parameter model that supports voice cloning from a few seconds of audio. To ensure high efficiency, the models are provided as optimized ONNX files with KV cache merging, enabling fast CPU-based synthesis.

Who it’s for

Developers and businesses looking to integrate lightweight, bilingual TTS capabilities into their applications, especially those targeting edge devices or those who want to avoid the expensive GPU requirements of larger models.

Highlights

  • Ultra-Lightweight: Core models are only 40M and 80M parameters, maintaining high sound quality (DNSMOS > 4.0).
  • Native Bilingualism: Seamlessly handles Chinese and English without needing to switch branches.
  • On-Device Friendly: Optimized for CPU and embedded devices via ONNX packaging.
  • Pronunciation Correction: Supports the use of Pinyin to manually correct mispronounced polyphonic characters or proper nouns.
  • Voice Cloning: The 80M model allows for quick voice cloning using short audio samples.

Related

  • Project
    kaanozhan/Frame

    Frame is an Electron‑based desktop IDE that adds a persistent, git‑tracked “project memory” for AI coding assistants (Claude Code, Codex, Gemini). It creates a `.frame/` folder with rules, architecture maps, task lists and a spec‑driven workflow, injects this context into each AI session, and anchors updates to git commits via pre‑commit hooks. Features include a multi‑terminal grid, task/spec panels, GitHub integration, fast file lookup, and multi‑AI switching. Intended for developers who want reliable, shareable context when working with LLM‑powered agents.

  • Project
    BakeLens/crust

    Crust is an open‑source, locally‑run proxy that sits between AI coding assistants and LLM providers. It inspects every tool call (file reads, shell commands, network requests) and blocks actions that could expose secrets or harm the host. The gateway works as an HTTP proxy, MCP/ACP stdio wrapper, or Docker container, and includes 42 built‑in security rules, 51 DLP patterns, and a plugin system for custom policies. It runs entirely on the user’s machine, stores logs encrypted in the OS keyring, and offers iOS integration via a Swift package.

  • Project
    agenvoy/Agenvoy

    Agenvoy is a Go‑based, single‑binary, self‑hosted AI‑agent platform that lets large language models act on your computer. It runs sandboxed tools (via bubblewrap), shares them through an MCP server, offers a web dashboard, voice interaction, scheduling, file‑based RAG, and private Telegram/Discord access.

  • Project
    composio-community/open-chatgpt-atlas

    Open ChatGPT Atlas is an open‑source Chrome/Edge extension (and optional Electron app) that lets an AI control the browser (via Gemini 2.5 Computer‑Use) and call over 500 third‑party services (via Composio). Install with Node, load the built `dist` folder, add your Google and optional Composio API keys, then use the sidebar chat to give natural‑language commands like “click this button” or “create a GitHub issue.” It offers visual feedback, safety confirmations, and runs entirely client‑side.