OpenAI has quietly stepped into hardware with the launch of the Codex Micro, a compact mechanical keypad built specifically for developers using its AI coding platform, Codex. Priced at $230 (roughly ₹20,000), the limited-edition device is available while stocks last, marking OpenAI's first commercially available hardware product.Unlike the consumer AI device that OpenAI is developing with former Apple design chief Jony Ive, the Codex Micro is aimed squarely at software developers. Built in collaboration with boutique keyboard maker Work Louder, the device features 13 programmable mechanical keys, a rotary dial, joystick, and a touch sensor designed to make interacting with multiple AI coding agents more tactile and efficient. Six illuminated keys display the real-time status of Codex agents, allowing developers to see whether a task is running, completed or requires attention without constantly switching back to the desktop application. Users can also assign shortcuts for actions such as push-to-talk, accepting or rejecting code changes, and launching workflows. The rotary dial can even adjust the reasoning level of an AI agent on the fly.The launch, however, has gone largely under the radar. Much of the attention around OpenAI hardware continues to revolve around its much-anticipated AI device with Jony Ive, making the Codex Micro appear more like a niche accessory than a headline product. Its limited production run, developer-focused audience and distribution through a specialist retailer rather than mainstream channels have also contributed to the subdued rollout.Despite the quiet debut, the Codex Micro offers a glimpse into how AI-native workflows could evolve. As developers increasingly supervise multiple AI agents working in parallel, OpenAI appears to be betting that dedicated physical controls can improve productivity in the same way specialised keyboards and macro pads have done for creative professionals. Whether it remains a niche tool or becomes the foundation for a broader hardware strategy will likely depend on how quickly AI-assisted software development becomes mainstream.