JP Morgan: Raised target price of KNOWLEDGE ATLAS (02513) to 1,400 Hong Kong dollars, downgraded MiniMax-W (00100) rating to "Neutral"

JPMorgan Chase released a research report on the Chinese artificial intelligence large model industry, pointing out that as the monetization path of large models converges towards enterprise workflows, API consumption, code, and intelligent agents, investors will increasingly focus on the actual capabilities of the models, task completion rate, and pricing power. Against the background of AI adoption demand still exceeding the supply of inference computing power, developers who quickly withdraw from premium pricing reflect that their model capabilities have not been recognized by the market. The firm is relatively bullish on KNOWLEDGE ATLAS (02513) in the sector but has downgraded the investment rating of MiniMax-W (00100). The firm reiterated a "hold" rating on KNOWLEDGE ATLAS. The firm is optimistic about the "Price Maker" behavior displayed by KNOWLEDGE ATLASAI, which doubled API prices this year while still maintaining continuous growth in business volume, fully validating the market's recognition of the value of its large models. The firm raised KNOWLEDGE ATLASAI's revenue forecast for the fiscal years 2026 to 2030 by 26% to 42%, lowered its adjusted net loss forecast, and significantly increased its target price from HK$950 to HK$1,400. In contrast, J.P. Morgan downgraded MiniMax's investment rating from "hold" to "neutral" and lowered its target price from HK$1,100 to HK$400. This is mainly due to MiniMax pricing its new M3 model at twice the price of the previous generation M2.7 on June 8, but then permanently reducing the price by 50% just a week later, essentially falling back to a level comparable to M2.7, reflecting a decrease in visibility of its leading position in large models. To reflect the negative impact of the price reduction on profit margins, the firm lowered MiniMax's revenue forecast for the fiscal years 2028 to 2030 by 5% to 21% and expanded the adjusted net loss forecast for the fiscal years 2026 to 2028.