Zhou Hongyi: No Unlimited Token Plans Ever, AI Costs Will Only Rise with OpenClaw Example

Zhou Hongyi: No Unlimited Token Plans Ever, AI Costs Will Only Rise with OpenClaw Example

N
News Editor 01
2026-07-22 14:15:13
Qihoo 360 founder Zhou Hongyi stated at the 2026 Zhongguancun Forum that tokens can never be offered as unlimited monthly plans like mobile data. AI runs on compute power; the more complex the task, the higher the cost. Heavy OpenClaw users already pay up to 30,000 yuan monthly.
Zhou HongyiToken EconomyAI ComputeOpenClawZhongguancun Forum

Zhou Hongyi, founder of Qihoo 360, directly dismissed market hopes at the 2026 Zhongguancun Forum annual conference: tokens will never be available as unlimited monthly subscriptions like mobile data. His answer was a flat "impossible."

Token vs. Data: Computing Cost Flips the Model

Zhou explained that traditional internet runs on data traffic, where fiber bandwidth is almost infinite and marginal costs drop as users increase, enabling unlimited monthly plans. But AI runs on compute power and intelligence.Token is the unit measuring AI compute consumption, with a relatively fixed unit price; the more you use, the more you pay. This is the opposite of the declining marginal cost model of mobile data.

Real numbers back him up. Heavy OpenClaw users consume about 100 million tokens monthly, costing roughly 7,000 yuan. Cheetah Mobile CEO Fu Sheng disclosed that a full-configuration OpenClaw subscription approaches 30,000 yuan per month.Even with no tasks assigned, OpenClaw's built-in heartbeat mechanism consumes about $20 daily. One Shenzhen engineer lost $12,000 in three days after his API key was stolen, fueling "token anxiety" among users.

OpenClaw Strategy: Free Software, Paid Tokens

Zhou specifically highlighted OpenClaw's model — free software, paid tokens — as a way to cultivate user payment habits. He believes Alibaba's decision to establish a Token Hub business group confirms the direction of the token economy. As OpenClaw surges in China, inference compute demand is about to explode.Zhou predicted inference compute needs will multiply by hundreds or even thousands of times, distinguishing training compute (gateway for big companies) from inference compute (the real explosion point ahead).

This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan.
400

Disclaimer:

The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.

Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.