P95 latency drops by 91%. What does that mean? It used to take ten seconds, now it's done in the blink of an eye.

Mem0 announces the core long-term memory algorithm research: extracting key facts through a two-stage pipeline and updating memory to avoid forgetfulness. Under the LOCOMO benchmark, accuracy is 26% higher than OpenAI's built-in memory, P95 reasoning latency is reduced by 91%, and token consumption is decreased by 90%. The enhanced variant Mem0ᵍ introduces a graph database to capture cross-session entity relationships. From memory retrieval to response in production only takes 0.71 seconds, far better than nearly 10 seconds of full context. The research has been accepted by ECAI, and the code has been open-sourced on GitHub.