Xiaomi (小米) Didn't Cut Prices This Time Either

Consumers expected cuts. Chatter moved to AI billing instead.

Xiaomi (小米) held prices steady while many shoppers hoped for reductions—again. Why did consumers walk away empty-handed? Because the online conversation was dominated not by handset promos but by a separate flap over AI model access and third‑party billing, it has been reported that users on Chinese platforms focused more on the unexpected costs of using advanced models through proxies and agent tools than on smartphone discounts.

Third‑party access to Claude Code now pay‑per‑token, reportedly

Boris Cherny, head of Claude Code, reportedly confirmed that continued access via third‑party channels will require extra payment: token‑based billing rather than a flat subscription, with no packaged plans though discounts may apply. It has been reported that autonomous proxy tools such as “OpenClaw” can run 24/7 at machine speed, driving a subscription user’s effective consumption into levels that, under API billing, could reach thousands or even tens of thousands of dollars. The same social posts also alleged that a figure nicknamed “Peter” joined OpenAI and that this played a role in services being blocked—claims that remain unverified and should be treated with caution.

Why this matters beyond price tags

For Western readers: China’s consumer tech market is increasingly intertwined with AI tooling and developer ecosystems. Subscription pricing assumes human‑scale use. But automated agents convert that into industrial‑scale throughput, exposing a gap between product plans and real‑world use. There is also a geopolitical backdrop—export controls and international competition over AI infrastructure mean that access, pricing and enforcement are sensitive issues for both Chinese firms and foreign providers.

The takeaway

Xiaomi’s decision not to cut prices is only one part of a larger story about cost, access and control in China’s tech ecosystem. Consumers waiting for cheaper phones aren’t the only ones budgeting carefully; developers and hobbyists now face a new calculus when choosing between subscriptions and pay‑per‑token APIs. Who wins when consumption moves from people to machines? The answer will shape both pricing and policy in the months ahead.