MemoraX AI Ranks #1 on Agent Memory Leaderboard, Signaling a New Phase for Long-Term AI Memory
SAN FRANCISCO, Aug. 17, 2026 (GLOBE NEWSWIRE) -- As AI agents move beyond one-off conversations toward longer-term
Press Release Disclaimer: This is a press release distributed through the XPR Media network. It has not been independently verified by our newsroom.
SAN FRANCISCO, Aug. 17, 2026 (GLOBE NEWSWIRE) — As AI agents move beyond one-off conversations toward longer-term interaction and collaboration, a new question is emerging: Can AI really learn from what it has done before?
MemoraX AI believes the answer starts with memory. The AI technology company today announced that it has ranked #1 on the inaugural Agent Memory Leaderboard (AML), taking the top position in the commercial product category and leading across all seven evaluated memory capabilities.
Recently unveiled, AML aims to establish a more consistent way to evaluate and compare AI agent memory. In its inaugural commercial ranking, MemoraX placed ahead of international memory platforms including Mem0, Vectorize and Supermemory, as well as APIs submitted by major technology companies including Tencent and NetEase, scoring 58.02 overall.
Why AI Memory Matters
An AI agent may successfully solve a difficult problem today, but when a similar problem appears tomorrow, it can still behave as if it has never seen it before.
It may repeat the same exploration, make the same mistakes or ask users to explain the same preferences and background again. A larger context window does not necessarily solve the problem. More history can also bring irrelevant information, outdated assumptions and failed attempts. The harder problem is knowing what is worth remembering, when to recall it and which past experience can still be trusted.
As AI agents move from answering individual prompts toward longer and more continuous tasks, memory is becoming an increasingly important part of how they operate.
Beyond Storing and Retrieving
MemoraX is building long-term memory technology that helps AI systems retain experience, maintain continuity and learn from what happens over time.
Rather than simply storing and retrieving past information, MemoraX focuses on the harder problem — helping AI determine what to keep, when to recall and how to adapt past experience to new contexts. Its technology combines learnable memory strategies, endogenous memory and a self-evolving Agent Harness, creating a continuous feedback loop between memory, tasks and outcomes. The goal is not to make AI remember everything. It is to help AI build on useful experience instead of starting over.
“AI agents are becoming increasingly capable, but without continuity, their capabilities are still limited,” said Jianye Hao, founder of MemoraX AI.“Memory allows an agent to build on previous experience instead of starting over each time. We are proud of this result, but we see it as the beginning of a much larger opportunity.”
A Milestone for MemoraX
The inaugural AML ranking provides a new reference point for a field that has historically been difficult to evaluate independently. For MemoraX, ranking first in the commercial category and across all seven evaluated capabilities represents an important validation of its approach to long-term AI memory.
The result comes as MemoraX recently completed a Seed++ financing round of tens of millions of U.S. dollars. The funding will support continued research and development in long-term AI memory and the development of related products.
As AI systems take on longer and more complex tasks, MemoraX believes memory will become an increasingly important layer of the AI stack, helping AI not only respond, but remember, learn and continue.
About MemoraX AI
MemoraX AI is building the memory layer for long-horizon intelligence. Its technology helps AI systems retain experience, maintain continuity and build on what they have learned over time. Built on learnable memory strategies, endogenous memory and a self-evolving Agent Harness, MemoraX is developing memory systems for AI agents and other long-term intelligent applications. Learn more at: https://x.com/MemoraX_AI.
Rosanne: biz@memorax.net
