Haru-4B: the open AI voice model that talks and listens at once
Haru-4B from Japan's higurashi is a free, open voice model that listens and speaks at the same time, in 9 languages. What it does and how to run it.
Checked and updated 16 November 2026 at 03:45 UTC

Key points
- Haru-4B is a free, open-weights "full-duplex" voice model from higurashi that listens and speaks at the same time, like a phone call, instead of waiting for you to finish.
- It speaks 9 languages, takes a text card for its character and a 3-second to 3-minute clip for its voice, and is released under Apache 2.0.
- It's a research base model: it runs on an NVIDIA GPU with about 12 GB of memory or as a native Mac app on Apple silicon, but it isn't a reliable source of facts.
Haru-4B is a free, open voice AI model that listens and speaks at the same time, so a conversation with it feels like a phone call rather than a walkie-talkie. Japanese company higurashi released it on November 12 under the Apache 2.0 license, and it runs on an NVIDIA GPU with about 12 GB of memory or as a native app on Apple silicon Macs.
It's been climbing Hugging Face's trending list this week. Here's what's new about it and what it can and can't do.
What makes Haru-4B different
Most voice assistants take turns: you talk, they wait for silence, then they answer. Haru is full-duplex. From the model card:
- Listens while it talks. Every 133 milliseconds it takes one slice of your audio and produces one slice of its own audio and text, so it can react mid-sentence.
- Fast on a laptop. higurashi's technical blog measured about 65 ms per step on a MacBook with an M1 Max (int8 weights), plus up to 133 ms of built-in framing delay.
- Nine languages: English, Spanish, Portuguese, German, French, Chinese, Japanese, Russian and Korean.
- You set the character and the voice. A text "character card" sets who it is (name, persona, situation). A short clip, 3 seconds to 3 minutes of one speaker, sets how it sounds.
How it was built
| Haru-4B | |
|---|---|
| Base language model | Qwen3.5-4B (32 layers) |
| Audio codec | VibeVoice acoustic tokenizer, 24 kHz |
| Training audio | 40M+ hours of speech in 160+ languages |
| Dialogue training | 13M+ hours of separated two-person conversations, up to 60 minutes long |
| Training compute | 256 NVIDIA B200 GPUs for about 8 days |
| License | Apache 2.0 |
The blog says LLM- and TTS-generated conversations were left out of the training data, so it learned from real human dialogue. Sources: model card and blog.
How to try it
From the GitHub repository:
- NVIDIA GPU (about 12 GB): clone the repo, run
pip install -e ./haru, thenpython -m haru.server --hf-repo higurashi-ai/Haru-4Band openhttp://localhost:8998in your browser. - Pick a character and give it a voice. Haru ships with no voice clips, so upload or record one.
- Mac with Apple silicon: the repo has an MLX version with a native macOS app you build with a script (
build_app.sh), then openHaru.app. - Use wired headphones and a wired mic. Bluetooth adds 100 ms or more of lag.
Running models at home keeps your audio on your own machine; our guide to on-device vs cloud AI explains why that matters.
Limits to know
- It's a base model for research. higurashi says it holds a natural conversation but isn't a reliable source of facts and doesn't follow instructions the way an assistant does.
- Voice imitation. It copies the voice in your reference clip. Only use voices you have the right to use.
- Small brain. The blog says the 4B backbone limits answer quality, and the team plans to scale to 100B+ parameters. A detailed technical report and training code aren't out yet.
If you want a text-to-speech model rather than a conversation partner, see ZONOS3-Mini. Haru is built on a small Qwen model; for the bigger open Qwen, see Qwen3.8-27B.
AI Tools This Week is not affiliated with higurashi.
Frequently asked questions
Is Haru-4B free?
Yes. The weights are on Hugging Face under the Apache 2.0 license and the code is on GitHub. You need your own hardware: an NVIDIA GPU with about 12 GB of memory, or a Mac with Apple silicon for the native app.
What does full-duplex mean?
The model hears you and talks at the same time. Every 133 milliseconds it takes in a slice of your audio and produces a slice of its own, so it can react while you're still talking, the way people do on a phone call, instead of waiting for you to stop.
Can Haru-4B clone a voice?
It imitates the voice in a reference clip you give it (one speaker, 3 seconds to 3 minutes of clean speech). It ships with no voices, and higurashi says to only use voices you have the right to use.
Sources
Related
Idol image-to-3D model: what it is and is it free?
Speridlabs' Idol turns one photo into a 3D Gaussian splat, including the sides you can't see. Free under Apache 2.0, but it needs an NVIDIA GPU.
Qwen3.8-27B: what it is, is it free and can you run it?
Alibaba's Qwen3.8-27B is a free, open AI model that runs on a strong home PC. What it does, file sizes, how to install it and benchmarks.
ZONOS3-Mini: what it is and is it free? (Nov 2026)
Zyphra's ZONOS3-Mini is a free, open-source text-to-speech and voice cloning model that runs in your browser. Features, languages, limits.