ReportGem ReportGem 中文

GLOBAL RESEARCH ARCHIVE

China Artificial Intelligence Capital accelerates race; Zhipu (OW) PT to HK$2400 on higher ARR visibility; MiniMax (N) needs stronger model validation

Published: 2026-07-12Institution: JPMorganCompany / ticker: 0100.HK,2513.HKPages: 19Original language: 英语Evidence page: 2

Research evidence excerpt

China Artificial Intelligence Capital accelerates race; Zhipu (OW) PT to HK$2400 on higher ARR visibility; MiniMax (N) needs stronger model validation

Olivia Xu Asia Pacific Equity Research

(86-21) 6106 6138 12 July 2026

olivia.w.xu@jpmorgan.com

Equity Ratings and Price Targets

Mkt Cap Price Rating Price Target

Company Ticker ($ mn) CCY Price Cur Prev Cur End Prev End Date

Date

Zhipu AI 2513 HK 46,320 HKD 1,640.00 OW n/c 2,400.00 Dec-26 2,000.00 n/c

MiniMax Group Inc - H 100 HK 7,821 HKD 268.60 N n/c 240.00 Dec-26 300.00 n/c

Source: Company data, Bloomberg Finance L.P., J.P. Morgan estimates. n/c = no change. All prices as of 10 Jul 26.

China AI foundation-model industry: capital accelerates the

race, while execution determines outcomes

The foundation-model industry is becoming structurally more capital intensive:

maintaining competitiveness requires continuous investment across pre-training,

post-training, reinforcement learning, evaluation infrastructure and inference

deployment. Unlike traditional software businesses where incremental distribution

can scale at relatively low marginal cost, frontier AI models require repeated

investment cycles to improve capability and support growing usage. This dynamic is

increasingly visible among China’s independent LLM providers –since the start of

2026, Chinese independent LLM providers have collectively raised or announced

over US$20+bn via IPOs, placements and private rounds.

However, the increasing importance of capital does not mean funding alone

determines competitive outcomes, as model development remains dependent on

several other factors, including the ability to identify the right technical direction,

attract high-quality research talent, build efficient engineering processes and organize

resources effectively.

The English excerpt is extracted automatically from the cited source page and may contain layout or recognition errors. It is never batch translated.

Open report viewer