Back to Arena

AnythingLLM

by Mintplex Labs

System Card

OrganizationMintplex Labs
Released2023-06
Architecturevector-rag / Full-stack ChatGPT clone with workspace memory
DetailsMonorepo (Vite/React + Node Express + collector + embeddable widget). Workspaces encapsulate document sets and chat history; supports many LLMs + vector DBs out of the box.
Parameters—
Domainrag-retrievalagent-memory
Open SourceYes
WebsiteVisit
CodeRepository
workspacesbrowser-extensionmulti-useragents

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
1/2
LoCoMo
57.17p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
59.19p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
60.34p
HotpotQA
61.116p
Agent Task Memory
1/1
AgentBench-Mem
59.95p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:AnythingLLM (Mintplex-Labs/anything-llm); evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)AnythingLLM (Mintplex-Labs/anything-llm); evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)AnythingLLM (Mintplex-Labs/anything-llm); evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)AnythingLLM (Mintplex-Labs/anything-llm); evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)AnythingLLM (Mintplex-Labs/anything-llm); evaluated on LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding (Tsinghua KEG, 2308)AnythingLLM (Mintplex-Labs/anything-llm); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)