Haru-4B: A voice AI that can listen and speak at the same time Higurashi AI released Haru-4B, a pretrained full-duplex spoken dialogue foundation model that listens and speaks simultaneously with about 65 ms end-to-end latency on a MacBook. Haru-4B runs at 7.5 Hz native full-duplex speech-to-speech, processing incoming and outgoing audio every 133 ms, and was pretrained on 40 million hours of real speech spanning over 160 languages, with primary focus on nine including English, Spanish, Chinese, Japanese, and Korean. The model, built on Qwen3.5-4B, uses 64-dimensional continuous acoustic latents and a lightweight MLP head with one-step MeanFlow sampling instead of an acoustic Transformer, with weights on Hugging Face and code on GitHub for further fine-tuning and research. Haru-4B: A Low-Latency and Multilingual Full-Duplex Continuous Spoken Dialogue Foundation Model We introduce Haru-4B, a full-duplex continuous spoken dialogue foundation model capable of listening and speaking simultaneously. Haru-4B focuses on natural human-AI interaction, learning conversational behaviors such as backchannels, interruptions, and turn-taking from large-scale real-world speech, rather than relying on synthetic conversations or explicitly scripted turn-taking behavior. Our goal is to make interacting with AI feel more like a natural human conversation, unlike interacting with agentic assistants. We are releasing Haru-4B as a pretrained base model , providing a foundation for further fine-tuning and research on real-time spoken interaction. The model weights are available on Hugging Face https://huggingface.co/higurashi-ai/Haru-4B , and the code is available on GitHub https://github.com/higurashi-ai/Haru . Haru-4B at a Glance - Ultra-Low-Latency Interaction: About 65 ms end-to-end latency on a MacBook. - 7.5 Hz Native Full-Duplex Speech-to-Speech: Simultaneous processing of incoming and outgoing speech every 133 ms through synchronized audio-text streams. - Single-Step Continuous Audio Generation: A lightweight MLP head with one-step MeanFlow sampling, eliminating the need for an additional acoustic Transformer. - 40 Million Hours of Real Speech: Large-scale multilingual pretraining without incorporating LLM- or TTS-generated turn-based synthetic dialogues. - Multilingual Speech Capabilities: Training data spanning over 160 languages, with a primary focus on nine languages: English, Spanish, Portuguese, German, French, Chinese, Japanese, Russian, and Korean. - Long-Form Full-Duplex Training: Trained on full-duplex conversations up to 60 minutes long, enabling the model to learn extended conversational dynamics and long-range interaction patterns. - Controllable Character and Voice: Supports character customization through text prompts and reference audio, enabling control over persona, speaking style, and voice characteristics. - Open Pretrained Base Model: We are releasing model weights Hugging Face https://huggingface.co/higurashi-ai/Haru-4B , code GitHub https://github.com/higurashi-ai/Haru , and demos to support further research. 1. Architecture: Native Full-Duplex at 7.5 Hz Haru-4B is built on Qwen3.5-4B