OpenAI is stepping into hardware with a new device expected in 2027. The product is a rechargeable, portable speaker designed to feel humanlike. It won’t have a screen but will use cameras and sensors to understand its surroundings.
This speaker is more than just a voice assistant. It will act as a “humanlike AI companion” that talks naturally and controls smart home devices. Mechanical parts will move on their own, creating an illusion that the device is alive.
At the core of this speaker will be GPT-Live. This is an advanced voice AI model that can listen and speak at the same time. It processes information quickly and allows natural conversations where users can interrupt or pause.
OpenAI believes voice will soon become the primary way people interact with AI. The new voice models, GPT-Live-1 and GPT-Live-1 mini, are designed to make chats feel more natural and seamless. They will route questions to OpenAI’s latest text models behind the scenes.
Design and Development
OpenAI acquired a design company called io in 2025 for $6.5 billion. io was founded by Jony Ive, the former Apple designer. This acquisition brought in key creative talent to help shape the speaker’s look and feel.
OpenAI also hired over 400 former Apple employees. Paul Meade, a former Apple exec, now leads OpenAI’s hardware division. These moves show OpenAI’s serious push into hardware and design quality.
Legal Challenges and Release Timeline
Apple has filed a lawsuit against OpenAI. The suit accuses OpenAI of trade secret theft and claims io was complicit. Apple also named two former employees, Chang Liu and Tang Yew Tan, alleging they downloaded confidential hardware files.
Despite this, OpenAI plans to launch the speaker in 2027. The release could face delays depending on how the lawsuit unfolds. Meanwhile, OpenAI is teasing other hardware, like the Codex Micro, set to come out on July 15th, 2026.
The speaker will include a rechargeable battery, making it easy to carry from room to room. It will support smart home controls, media playback, answering questions, and messaging—all without a screen.
OpenAI has said inference costs have dropped enough to make voice AI economically viable. They will start rolling out the voice models to free users soon. Audio data is stored for 30 days, but users can delete or export it. Training on user queries is off by default unless users opt in.
Atty Eleti, an OpenAI spokesperson, said, “This is the beginning. Over time, we think this will also unlock the ability to use voice as kind of the primary interface to computing.”
OpenAI’s speaker could change how we interact with AI at home. It aims to feel alive, listen and talk like a person, and blend smart tech with natural conversation. The next few years will show if this vision becomes reality.
Based on
- OpenAI’s first device will reportedly be a ‘humanlike’ rechargeable speaker — engadget.com
- OpenAI’s first hardware device is reportedly a screenless speaker that can move | TechCrunch — techcrunch.com
- OpenAI may announce a ChatGPT smart speaker this year | The Verge — theverge.com
- OpenAI bets voice will become AI’s primary interface with new models — axios.com
- GPT-Live, OpenAI’s New Voice Model, Wants You to Talk Over It – Business Insider — businessinsider.com
