Morgan Stanley's Asia-Pacific research team has released a new industry report titled "China's AI Path: More Bang For The Buck," significantly raising price targets for two leading Chinese AI companies. The report expresses optimism regarding the continued realization of China's AI value proposition in terms of annual recurring revenue (ARR) and valuation. The price target for MINIMAX-WP(00100) has been increased from HK$990 to HK$1,100, while its industry rating is maintained at "In-Line." The report highlights MINIMAX-WP and other companies with full-stack AI capabilities as key recommendations. The report indicates that while Scaling Law remains valid and leading US large language models (LLMs) are advancing faster than expected, Chinese AI firms have not yet encountered significant bottlenecks in engineering efficiency scaling. Given that computing power remains a primary constraint, top-tier Chinese and US models are now nearly equivalent on the Artificial Analysis intelligence index—with MINIMAX-WP's M2.7, Zhipu's GLM-5.1, Moonshot's K2.6, and DeepSeek's V4 all scoring between 50 and 54 points. Morgan Stanley concludes that the AI capability gap between China and the US has narrowed to just 3-6 months. A key advantage for Chinese models is their cost efficiency: they achieve comparable intelligence levels at just 15-20% of the inference costs of their US counterparts. The report attributes gains in engineering efficiency to advancements in three main areas: architectural improvements involving dense models, mixture-of-experts (MoE), and attention mechanisms; post-training enhancements like reinforcement learning and model distillation; and inference infrastructure optimizations related to hardware and KV cache efficiency. The recently launched DeepSeek-V4 is cited as a prime example of this progress. Data from OpenRouter shows a historic shift in token consumption share. The portion of tokens used by leading Chinese models surged from 5% in April 2025 to 32% by March 2026. During the same period, the share for leading US models dropped significantly from 58% to 19%. Token usage for MINIMAX-WP, Zhipu, and Alibaba in February-March 2026 increased by 4 to 6 times compared to December levels, driven largely by breakthroughs in coding and agentic workflow capabilities. Contrary to market concerns about AI model commoditization, advanced models are demonstrating clear pricing power. Zhipu's prices have nearly doubled since the start of the year, and MINIMAX-WP's KV cache pricing has also doubled. Morgan Stanley explicitly anticipates that MINIMAX-WP will implement significant price increases following its M3 model upgrade, which could serve as a catalyst for ARR growth in the next phase. Regarding ARR projections, Morgan Stanley provides a base-to-optimistic forecast: leading model providers are expected to achieve ARR between $1 billion and $1.5 billion by the end of 2026, potentially climbing to $2.5 billion to $5 billion by the end of 2027. This represents an annualized growth rate of 3 to 5 times. Currently, MINIMAX-WP's ARR reached $150 million in February 2026, while Zhipu's ARR hit $250 million in March. In the Hong Kong stock market, Zhipu and MINIMAX-WP are highly likely to be included in the Hang Seng Tech Index on June 8, with a combined weighting of 5-7%. This inclusion is projected to attract between $1.25 billion and $1.75 billion in passive fund inflows. The report also notes that technology companies now account for 40% of Hong Kong IPO fundraising year-to-date and 43% of the pipeline, indicating that AI is becoming a long-term force reshaping the composition, liquidity, and capital flows of Hong Kong stock indices.