I tried out OpenAI’s new AI keypad — which will be fun for some coders and slightly mystifying to everyone else
TechCrunch Lucas Ropek ● Covered by 8 sources
OpenAI shipped a $230 keypad called Micro that pairs with ChatGPT and Codex. It's OpenAI's first hardware, but coders on Reddit mostly hate it.
Based on reporting by TechCrunch, Lucas Ropek — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
OpenAI's first hardware product isn't a phone or a wearable. It's a keypad. Called Micro, built with keyboard specialist Work Louder, it's a chunky little accessory with six customizable "agent" keys and six command keys designed to let you fire off ChatGPT and Codex tasks without touching your laptop's keyboard.
The timing is a little awkward. Apple sued OpenAI weeks ago over alleged trade secret theft, and rumors of a separate OpenAI smart-home device built by ex-Apple engineers have only added fuel to that fire. Against that backdrop, Micro arrives in packaging so clean and white it practically dares you to compare it to an iPhone box. TechCrunch's Lucas Ropek noticed the resemblance immediately, and so did plenty of people online.
Functionally, Micro lets you assign ChatGPT sessions or Codex commands to individual keys, so you can hop between projects without alt-tabbing through browser windows. There's a dedicated dictation button — hold it, talk, tap send, and your request goes straight to the AI. Keys light up in different colors depending on status: white for idle, blue for thinking, green for done, red for errors. Customization happens inside a Micro tab in ChatGPT itself, where you can tweak brightness and rebind commands.
The reception from the people Micro is actually built for has been rough. Reddit threads call it "a prank and not a real product," and reviewers elsewhere, including Aftermath, have hammered the $230 price tag as tough to justify against cheaper macropads already on the market. Ropek's own experience landed somewhere in the middle: confusing at first, genuinely satisfying once he'd mapped out his keys and gotten used to switching between sessions with a tap and a voice command.
But the core question nobody's really answered is why anyone needs a dedicated physical device to do what a keyboard shortcut or a browser tab already does. Micro's appeal collapses almost entirely into one variable — how much ChatGPT you use daily. For casual users, it's a novelty paperweight with nice buttons. For power users glued to Codex all day, it might shave off real friction. For everyone else, it's a $230 curiosity waiting for OpenAI's actual hardware ambitions to catch up.
My take — AI-written commentary, not fact-checked reporting
This is OpenAI testing the waters with a low-stakes toy while the real hardware fight — the one with Apple, the lawsuits, the ex-Apple engineers building something bigger — plays out behind closed doors. Micro isn't a product so much as a marketing exercise dressed up as one, and the fact that hardcore coders are trashing it tells you the target audience hasn't been won over yet. I'd rather OpenAI spend its hardware energy on something open and repairable than another walled-garden gadget chasing Apple's aesthetic without Apple's decade of hardware muscle behind it.
Read more about this at: TechCrunch