List your company

OS Task on Lifelong Agent Bench (val)

74.6Success RateMemRL
64.92867.43969.9572.461Jan 6, 2026

Evaluation results

MethodLinks
MemRLModel=GPT-4o-mini2026.0174.6
MemPModel=GPT-4o-mini2026.0172
RAGModel=GPT-4o-mini2026.0171.3
Mem0Model=GPT-4o-mini2026.0168.6
No MemoryModel=GPT-4o-mini2026.0167.3
Self-RAGModel=GPT-4o-mini2026.0165.3