OpenAI’s first hardware reveal! A small speaker that can “walk around” by itself, with a built-in camera, and it can ask and answer questions and talk as an AI companion

Bloomberg quoted sources on Tuesday (7/14) revealing exclusively that OpenAI’s first long-awaited consumer hardware is a smart speaker that “moves on its own” and has “no screen.” It has a rechargeable battery, a camera, and multiple sensors, allowing it to be carried casually from one room to another and understand the environment around the user. The device is positioned as an “AI companion,” with voice powered by a more advanced version of GPT-Live, enabling it to both listen and speak at the same time. OpenAI plans to announce it later this year, with a formal release in 2027.
(Background brief: OpenAI’s first leaked AI hardware product—a smart speaker that can recognize faces, observe, and help you shop, with the earliest possible launch in early 2027)
(Background addition: OpenAI’s shock move! Acquiring the legendary Apple product designer Jony Ive’s io company)

Key takeaways

• Bloomberg’s leak: OpenAI’s first hardware is a screenless smart speaker with a built-in camera and sensors, capable of moving between rooms
• The device is positioned as an “AI companion,” using a more advanced version of GPT-Live for voice that can both listen and speak, naturally picking up after interruptions
• Planned to appear later this year, with release in 2027; development led by former Apple hardware lead Tang Tan

Bloomberg reported on Tuesday (7/14), citing insiders, that OpenAI’s much-labored first consumer hardware has finally shown its outline, and the answer has surprised many. It is not a phone, and not glasses, but a smart speaker with no screen that can move around inside your home on its own. This speaker includes a rechargeable battery, and users can simply carry it from the living room to the kitchen, then into the bedroom; the body is fitted with a camera and various sensors to understand the surrounding environment and the visual scene in front of it.

Over the past few decades, Apple, Amazon, and Google have poured effort into putting screens into the living room, while OpenAI has gone the opposite way—removing the screen and leaving the camera. The device is meant to “see” you, rather than have you stare at it.

A camera, with no screen

Based on Bloomberg’s description, the speaker is positioned as an “AI companion,” able to control smart home appliances, play content, answer questions, help reply to messages, and directly call on ChatGPT capabilities. It will also access personal information such as email, gradually learning about the user, then proactively offering the information it believes you need.

What’s most intriguing is the selling point of a “mechanical structure that can move on its own.” This speaker is not just passively waiting; it can turn and reposition itself. OpenAI wants to make it feel more like a “living object,” not a cold, lifeless speaker. As for the voice side, it is handled by a more advanced version of GPT-Live. Its feature is that it can both listen and speak simultaneously; if you pause mid-sentence, or suddenly interject, it can naturally pick up the conversation instead of sounding clunky in question-and-answer like older voice assistants.

Built by a former Apple team, but stuck in a lawsuit

This hardware did not emerge out of thin air for OpenAI. It traces back to last year’s OpenAI acquisition of the hardware startup io Products for $6.5 billion. Ive’s design studio LoveFrom also continues to participate in building the entire product line. Bloomberg notes that this is one of about five products OpenAI has in development, led by former Apple hardware engineering lead Tang Tan and former industrial design lead Evans Hankey—a bona fide “former Apple dream team.”

On one side is a hardware ambition to hire the entire former Apple team, and on the other is Apple’s injunction lawsuit. Before the device even launches, it is already being caught in the shadow of legal warfare.

The dream team also brings trouble. Apple has sued OpenAI, accusing it of stealing hardware trade secrets and seeking an injunction. If the injunction is granted, the speaker’s launch timeline will likely change as well.

OpenAI plans to officially announce the device later this year, but it will not be released until 2027. Bloomberg also reminds readers that the product is still under development, and specifications and details may still be adjusted. From rumors of an “AI iPhone” to today’s moving, screenless smart speaker, OpenAI’s intent to bring ChatGPT into your living room is already clear. But how smoothly this step can go will likely depend first on how its lawsuit with Apple plays out. We’ll keep following it.

Frequently Asked Questions

What is OpenAI’s first hardware? When will it be released?

According to Bloomberg, OpenAI’s first consumer hardware is a screenless, movable smart speaker with a built-in camera and sensors, positioned as an “AI companion.” It is planned to be announced later this year, with a formal release in 2027, but the product is still in development and details may be adjusted.

What special features does this OpenAI AI speaker have?

It has no screen but includes a camera and sensors that can understand the surrounding environment; it has a mechanical structure that can move on its own, aiming to feel more like a “living object”; its voice uses a more advanced version of GPT-Live that can both listen and speak, naturally handling pauses and interruptions.