OpenAI is fast-tracking a ChatGPT phone for mass production starting in early 2027, according to supply chain analyst Ming-Chi Kuo, with combined 2027–2028 shipments projected to reach around 30 million units. The device would run a customized version of MediaTek's Dimensity 9600, the chip set to launch in fall 2026 as the successor to the Dimensity 9500. That puts a first-generation OpenAI hardware product in roughly the same shipment tier as a typical Samsung flagship line.
Kuo's note, surfaced May 5, 2026, reframes what OpenAI's hardware push actually looks like. The Jony Ive collaboration has been described as a screenless companion device, but Kuo's reporting points to a conventional smartphone form factor running on mainstream silicon — not an experimental wearable.
The customized Dimensity 9600 is the technical centerpiece. Kuo says the chip's headline spec is its image signal processor, with enhanced HDR meant to improve real-world visual sensing — the kind of capability a multimodal ChatGPT assistant would lean on for live camera input. The base 9600 is expected to power flagships from vendors like Vivo and Oppo, whose Vivo X300 Pro and Oppo Find X9 Pro currently ship with the Dimensity 9500.
Key facts
- 01Ming-Chi Kuo says OpenAI is fast-tracking a phone for mass production starting in early 2027.
- 02Combined 2027–2028 shipments could reach around 30 million units, near typical Samsung flagship volume.
- 03The device will run a customized MediaTek Dimensity 9600, the chip slated to launch in fall 2026.
- 04Kuo says the headline spec is an image signal processor with enhanced HDR for real-world visual sensing.
- 05The phone is expected to pair LPDDR6 memory, UFS 5.0 storage, and a dual-NPU architecture for AI workloads.
Beyond the SoC, Kuo describes LPDDR6 memory, UFS 5.0 storage, and a dual-NPU architecture designed to run different kinds of AI computation simultaneously — for example, language and vision models in parallel. That spec sheet reads like a phone built backwards from the workload: keep a language model resident in memory while a separate neural engine handles camera, speech, or sensor input without thrashing.
“Combined 2027–2028 shipments could reach around 30 million units, putting a first-generation OpenAI phone in the same volume tier as a Samsung flagship.”— Jaeden Schafer
The 30 million-unit target across two years is the most striking number in Kuo's note. Apple ships well over 200 million iPhones a year, and a Samsung Galaxy S flagship line typically lands in the 25–30 million range over its lifecycle. Hitting that bar on a debut device, with no carrier relationships, no retail footprint, and no installed base, would be without precedent for a software company entering hardware.
OpenAI's path to those volumes runs through ChatGPT itself. The company has previously said ChatGPT serves hundreds of millions of weekly users, and converting even a small share of that base into hardware buyers — particularly the Plus and Pro subscribers already paying $20 to $200 a month — would clear the 30 million bar comfortably. The phone is, in that reading, a distribution play for an AI subscription, not a standalone hardware bet.
MediaTek is the more interesting commercial winner if the timeline holds. Qualcomm has dominated premium Android silicon for years, and a marquee design win with OpenAI on a customized Dimensity part would mark a meaningful shift in who the AI-native phone vendors choose. The custom ISP work also suggests OpenAI is willing to invest in silicon-level differentiation rather than ship a reference design.
Software is the harder problem. OpenAI would need either a deep Android fork or a ground-up operating system, plus app compatibility, an app store strategy, and answers on messaging, payments, and the everyday utilities users expect. Kuo's note covers the bill of materials, not the platform — and the platform is where Microsoft, Amazon, and Facebook all stalled when they tried phones.
The Jony Ive question also doesn't go away. OpenAI's $6.5 billion acquisition of Ive's io hardware startup was framed around a new computing paradigm, not a rectangular slab. Either the phone and the Ive device are two separate products on different tracks, or the early-2027 phone is the more conventional first step before a more ambitious form factor lands later.
Kuo has a strong but imperfect track record on Apple supply chain reads, and his OpenAI sourcing is newer. Mass production timelines also slip routinely — early 2027 in a May 2026 note is the kind of date that frequently becomes late 2027 by the time tooling and certification finish. The 30 million-unit projection is a forecast, not an order book.
Still, the direction is clear. OpenAI is moving from a model and app company into a vertically integrated stack — silicon partnership, custom ISP, dedicated NPUs, and a shipment target that would make it a top-tier phone vendor on day one. Whether the device sells 30 million units or 3 million, the strategic message is that the next platform fight isn't happening inside someone else's operating system.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
