JPMorgan Raises Zhipu Target, Cuts MiniMax as AI Pricing Power Diverges

With AI inference demand still outstripping supply and developers competing on model performance, the ability to command premium API prices has become a key test of model quality and commercial potential. JPMorgan said benchmarks may be optimized for specific tests, while publicly available prices that remain in place over time provide a better gauge of genuine demand. The pricing gap between Zhipu and MiniMax therefore directly affects their investment ratings and valuations.

On June 15, JPMorgan maintained its Overweight rating on Zhipu and raised its price target by 47% to HK$1,400 from HK$950. Zhipu has doubled its API prices this year while usage has continued to grow. By contrast, MiniMax announced a permanent 50% price cut on June 8, about one week after launching M3. JPMorgan downgraded MiniMax to Neutral and cut its price target by 63.6% to HK$400 from HK$1,100.