OpenAI’s first physical hardware product is called Codex Micro, it costs $230, and it’s already being confused with the other OpenAI device — the one from the Jony Ive collaboration that’s at the center of the Apple lawsuit. They’re different products with different purposes. Let’s sort that out first.

What Codex Micro Actually Is

Codex Micro is a compact hardware device focused on AI-assisted coding and development tasks. At $230, it’s positioned as an accessible tool for developers rather than a mass-market consumer product. The Codex name connects directly to OpenAI’s Codex model family — the AI systems trained specifically for code generation and completion.

The device is essentially a dedicated inference endpoint with a physical interface, allowing developers to run AI coding assistance locally or in environments where cloud connectivity is limited. It’s niche by design.

The Distinction From the Jony Ive Device

The other OpenAI hardware project — developed in collaboration with Jony Ive — is a fundamentally different concept: a consumer AI device designed to be a phone alternative, meant to be worn or carried, with ambient AI assistance as its core proposition. That device is what triggered Apple’s lawsuit (alleging intellectual property issues related to design elements and potentially employment of former Apple engineers).

Codex Micro is none of that. It’s a developer tool. The timing of its launch alongside the Apple legal drama has created confusion that’s commercially unhelpful for OpenAI, but the products are genuinely distinct.

OpenAI’s Hardware Ambitions

What’s interesting about Codex Micro as a first hardware move: OpenAI is starting with a professional-focused product rather than going straight for the consumer market. That’s a different approach than the Jony Ive device suggests for the longer term.

The $230 price point is accessible enough for individual developers but signals a specific professional use case rather than a mass-market push. Whether this becomes a meaningful hardware product category for OpenAI — or whether it’s primarily a demonstration of capability before the larger consumer play — is the more interesting question.