Back to Arena

CrewAI

by CrewAI Inc. (Joao Moura)

System Card

OrganizationCrewAI Inc. (Joao Moura)
Released2024-01
Architectureagentic-workflow / Crew-level short / long / entity / contextual memory
DetailsMulti-agent "crew" framework where each crew can enable short-term (task), long-term (cross-session), entity, and contextual memories. Backed by ChromaDB by default; Pydantic state for flows.
Parameters—
Domainagent-memory
Open SourceYes
WebsiteVisit
CodeRepository
crewchromadbshort-long-entitycertified-devs

Capability Profile

Benchmark Scores

5 of 14 benchmarks
Long-Context Retrieval
0/5
RULER
no data
NIAH
no data
LooGLE
no data
LongBench
no data
∞Bench
no data
Multi-Turn Recall
1/2
LoCoMo
79.381p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
76.856p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
73.263p
HotpotQA
77.587p
Agent Task Memory
1/1
AgentBench-Mem
7226p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:CrewAI (joaomdmoura/crewAI); evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)CrewAI (joaomdmoura/crewAI); evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)CrewAI (joaomdmoura/crewAI); evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)CrewAI (joaomdmoura/crewAI); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)CrewAI (joaomdmoura/crewAI); evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)