Xteink: phone-back e-paper bet nets ¥100M+ across five rounds

Xteink makes ultra-portable magnetic e-paper readers (X3, X4, X4 Pro) that stick to the back of a phone as a low-distraction screen for comics, light novels and short reads, with plans to add AI-agent-served information.

That the phone's idle back is the next attention entrance: a ~70g magnetic e-paper 'second screen' served by AI agents, not another standalone app-filled device.

Hu Yufei, a serial operator (Uber China city GM, Bluegogo co-founder, Ele.me crowdsourcing GM, OYO chief development officer, Neolix COO) and heavy e-paper user, saw that phones were eating fragmented reading time while standalone readers were too bulky to carry. After Chinese New Year 2025 he posted the product brief on Xiaohongshu — ultra-thin, magnetic, light, long battery — and early users co-decided details like screen bezel, page-turn button and hold feel. The first device shipped in Q4 2025.

Rounds were small but frequent: by April 2026 Xteink disclosed five angel-to-A rounds totalling over ¥100 million from Boyu Capital, Qingliu Capital, Xiyang Capital, Xiaohongshu, Matrix Partners China and Shunwei Capital, with most investors topping up. Official figures showed monthly sales growing 10x within six months of launch, overseas first-month sales above 10,000 units, and monthly revenue exceeding the whole of 2025 by April 2026. In August 2026 existing backers Shunwei, Matrix, Boyu and Xiaohongshu invested close to ¥50 million more in a rolling A-round closing, and GeekPark reported monthly sales at the hundreds-of-thousands level. The company is building a three-tier architecture (e-paper device + phone compute + cloud agents) and says a new product with agent-driven personalised feeds arrives in H2 2026.

Still scaling as of the latest sources (2026-08-10): Xteink runs X3/X4/X4 Pro lines, claims hundreds-of-thousands-level monthly sales, and raised its ~¥50M A+ extension in August 2026. The open test it named itself is whether users keep paying for content and software once the low-margin hardware is only 'a ticket'.

Positioning: the back of the phone is high-frequency real estate every competitor ignored, and attaching to it removes the 'carry another device' objection that keeps readers at home.

Public co-creation: a Xiaohongshu 'cloud shareholder' process validated the design and produced first-batch demand without paid promotion.

Subtraction as differentiation: a closed system with no app store, feeds or notifications stands out in an AI-hardware era of adding sensors and models.

Operator founder with capital efficiency: Hu's history of scaling networks made investors comfortable backing a tiny hardware team that says it refuses to buy growth.

AI as a second act: the three-tier device-plus-phone-plus-cloud architecture gives the screen a reason to stay on beyond reading.

Pick a surface others ignore (the phone's back) and let users co-create the product in public; the follow-on test is whether 'less' still sells once AI and content have to add function.

As of 2026-08-10, Xteink is scaling: X3, X4 and X4 Pro are on sale, GeekPark reports monthly sales in the hundreds of thousands (数十万), and an A+ extension of close to ¥50 million closed in August 2026 with existing backers. 36kr had already reported monthly revenue above the company's full-year 2025 figure by April 2026. Management is testing cloud/phone/device agent architecture, with a new product due in H2 2026. The open questions: whether low hardware margins convert into content and software revenue, and whether an early specification mislabeling dispute damaged community trust.