LeituraPesquisaRadarFramework de investimento
Entrar / Cadastrar
← Voltar para a leitura
NotíciasTópico · Haru 4BMaterial1 min de leituraEditorial DisconfirmAI

Japanese Startup higurashi Open-Sources Full-Duplex Voice Model Haru 4B

higurashi releases Haru 4B, a 4B-parameter full-duplex speech model under Apache 2.0, aiming to solve latency and non-verbal cue loss in traditional cascaded voice systems.

Verificado 2026-11-13 20:55 GMT+8

Japanese AI startup higurashi released Haru 4B, a full-duplex speech dialogue model, on November 13, 2026. The company has made the pre-trained weights and inference code available on Hugging Face under the Apache 2.0 license. With 4 billion parameters, the model is specifically designed to address the needs for naturalness and low latency in real-time voice interactions.

Unlike traditional "cascaded" architectures that sequentially process Automatic Speech Recognition (ASR), Large Language Models (LLM), and Text-to-Speech (TTS), Haru 4B employs an end-to-end full-duplex LSLM (Large Speech Language Model) technology. This architecture allows the model to simultaneously process three data streams—user voice, model voice, and text—synchronizing information approximately every 133 milliseconds. This enables the model to "listen while speaking," facilitating low-latency responses to user interruptions, backchanneling, and topic shifts, while preserving paralinguistic features such as tone and emotion.

According to higurashi's official announcement, Haru 4B was pre-trained on 40 million hours of multilingual data including Japanese and English, currently supporting nine languages. The model also incorporates the VibeVoice continuous speech tokenizer to balance audio quality with computational efficiency. Despite its fluent conversational capabilities, higurashi explicitly notes that due to its 4B parameter size, the model's logical reasoning ability is roughly equivalent to GPT-2 levels, highlighting that its primary value lies in its interaction mechanism rather than complex task processing.

Fontes

Entrar / Cadastrar
LeituraPesquisaRadarFramework de investimento
higurashi.aihigurashi↗
  • pc.watch.impress.co.jpPC Watch↗
  • Tópico

    Haru 4B

    日本初创公司higurashi于2026年11月13日开源了全双工语音对话模型Haru 4B。该模型采用Apache 2.0协议,基于Qwen3.5-4B架构,通过同步处理用户音频、模型音频和文本三个流,实现了“边听边说”的自然交互能力,支…

    2 etapas1 peçasAtualizado 2026-11-17 20:27

    Como chegamos aqui

    1. 2026-11-13higurashi发布Haru 4B,在Hugging Face开源模型权重,GitHub开源推理代码,并公布技术博客。higurashi.ai ↗
    2. 2026-11-15第三方开发者sakasegawa发布Haru-4B的GGUF量化版本,适配speech.cpp框架,提供不同精度下的性能测试数据。huggingface.co ↗

    O que ainda está indefinido

    • Haru 4B作为研究用基础模型,其在实际生产环境中的稳定性和安全性如何评估?
    • 社区开发的量化版本(如GGUF)在低精度下对自然对话流畅度的具体影响有多大?
    • higurashi计划何时发布更大规模的模型以缩小与‘人类般AI’愿景的差距?
    Ler o tópico completo →
    Continuar pesquisando →