Back to Arena

Mem AI

by Mem Labs

System Card

OrganizationMem Labs
Released2021-01
Architecturevector-rag / Semantic Note Graph (no-folder AI second brain)
DetailsMem builds a semantic map of all user notes without folders or taxonomies, using an AI engine that reads everything written and makes it retrievable via natural language. Mem 2.0 introduced Agentic Chat that can create, edit, and organize notes via direct actions, plus Heads Up which proactively resurfaces relevant notes at the right moment. The Copilot sidebar provides live related-note suggestions as you type.
Parameters—
Domainrag-retrievalpersonalization
Open SourceNo
WebsiteVisit
note-takingsecond-brainsemantic-searchagentic-chatknowledge-managementno-folders

Capability Profile

Benchmark Scores

6 of 14 benchmarks
Long-Context Retrieval
1/5
RULER
no data
NIAH
no data
LooGLE
no data
LongBench
603p
∞Bench
no data
Multi-Turn Recall
2/2
LoCoMo
56.36p
MemoryBank
59.22p
Cross-Session Memory
1/1
LongMemEval
55.52p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
64.923p
HotpotQA
59.19p
Agent Task Memory
0/1
AgentBench-Mem
no data
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:Mem AI vendor documentation; evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)Mem AI vendor documentation; evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)Mem AI vendor documentation; evaluated on LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding (Tsinghua KEG, 2308)Mem AI vendor documentation; evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)Mem AI vendor documentation; evaluated on MemoryBank: Enhancing LLMs with Long-Term Memory (Sun Yat-sen University, 2305)Mem AI vendor documentation; evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)