According to media reports, enterprise AI solution provider
Technical Moat: "Memory Science" Empowers Multimodal Large Models
Founded in April 2024,
Scenario Memory Ability: It aims to enable AI to possess human-like "scenario memory," improving AI's perception and decision-making accuracy in complex environments through structured storage and retrieval of past experiences.
Physical World Interaction: Unlike pure text-based dialogue models, Hongxiong AI focuses on multimodal technology, aiming to enable AI to perform more precise actions and feedback in the physical world.
Capital Momentum: Valuation Set to Double, Clear IPO Timeline
After this round of financing, the post-money valuation of
Series A+ Prospects: The company is currently in discussions for a Series A+ round, expected to be completed between July and August this year, with the valuation potentially doubling to 3 billion yuan.
IPO Roadmap: Founder and CEO Wen Deliang revealed that the company plans to submit its prospectus to the Hong Kong Stock Exchange in the second or third quarter of 2027, and intends to officially launch the IPO process in 2028.
Industry Insight: "Memory Completion" in the Era of Embodied Intelligence
In the second half of the large model race, how to prevent models from "forgetting immediately after turning" is key to achieving artificial general intelligence (AGI). Hongxiong AI enters the enterprise market through "memory science," not only solving the pain points of AI in long-term sequential tasks, but also providing new solutions for physical scenarios requiring continuous learning, such as industry and logistics.
Conclusion: From Perception to Memory
The rapid valuation increase reflects the market's growing demand for "memory-enhanced AI." As the Series A+ financing approaches, whether Hongxiong AI can ring the bell as scheduled in 2028 will become an important case study for observing the maturation of domestic AI startups.

